Beruflich Dokumente
Kultur Dokumente
Preservatives
Introduction
Many of the materials used in the manufacture of toilet preparations are
susceptible to biological degradation by micro-organisms. In this chapter
methods of combating biodeterioration of products will be discussed.
In the past the contribution of the microbiologist to the cosmetics industry was
considered of minimal importance.' More recently there has been an increasing
awareness of the problems of microbial spoilage of cosmetic and toiletry
products. Formerly the emphasis was placed on loss of aesthetic appeal of the
product with a consequential loss of profitability, while the health hazard to the
consumer was largely ignored. Since micro-organisms are ubiquitous, the body is
continually exposed to them and it therefore seemed unnecessary for toilet
preparations to be sterile. Although it was accepted that products should not
constitute any greater microbial hazard than that presented by the normal
environment, little was done to show that this was achieved.
Normally healthy individuals have considerable resistance to infection by
bacteria and fungi commonly found on their skin and in their usual environment,
but in susceptible individuals, for example the newborn, the very old, those in ill
health or under drug therapy, there is an increased probability of infection
developing. It should be remembered that a product may contain a growing
bacterial population even though there is no visible evidence of this. Such a
product placed in intimate contact with the skin, particularly if it is' broken or
damaged, may give rise to infection.' Examples in the literature which demon-
strate the importance of considering the possible health hazard of a contami-
nated product will be discussed later.
Preservatives are, therefore, added to products for two reasons: firstly to
prevent spoilage, that is, to prolong the shelf-life of the product, and secondly to
.3
protect the consumer from the possibility of infection It is recognized that
products may require protection from contamination during manufacture,
although preservation must never be used to hide bad manufacturing
procedures .4 Duke quotes examples illustrating how products containing
preservatives can be microbiologically unacceptable because of bad manufactur-
ing practice. It is also recognized that cosmetic products, possibly more so than
pharmaceutical products, are liable to consumer abuse. Although products
cannot be protected against extremes of abuse, such as the use of saliva in the
application of-eye make-up, the manufacturer should anticipate misuse when the
product is formulated .3
It is therefore imperative that the microbiologist be involved in the formula-
tion of a product from its development stage onwards. 5 It can no longer be
676 Harry's Cosmeticology
assumed that a manufacturer has fulfilled his obligation to the consumer if the
product in question leaves his premises in a microbiologically satisfactory
condition. If, however, products are to be expected to remain resistant to the
introduction of extraneous micro-organisms during their period of use, this will
impose a greater strain on the preservative system chosen and will make it even
more important for manufacturers to carry Out thorough and realistic tests both
during the development of new products and in the production and quality
control of existing lines.
The Council of the British Society of Cosmetic Chemists has published a
monograph on the hygienic manufacture and preservation of toiletries and
cosmetics. 6 Further indications of the importance currently attached to ensuring
the adequate preservation of all toilet preparations and cosmetics are evident in
the number of investigations into contaminated preparations and the possible
clinical significance of contamination. Some of these reports are discussed later
in the chapter.
Microbial Metabolism
Micro-organisms grow and multiply by utilizing the materials in their immediate
environment. In considering spoilage problems created by micro-organisms, the
variety of chemical reactions they can carry out and the rate at which-these can
occur must be taken into account. -
Bacteria and fungi are widely distributed in nature and there are few places on
or near the surface of the earth that are free from them. They occur, for
example, in such unlikely places as hot mineral springs, the effluents from gas
works, stagnant salt lakes and even in the essentially anhydrous environment of
diesel fuel. The only places where micro-organisms are not found are those in
which a sterilizing influence prevails or in the interior tissues of healthy animals
and plants. In growing, bacteria and fungi can cause rapid and profound changes
in their immediate environment and in the synthesis of new protoplasm many
complex chemical reactions are accomplished within a remarkably short period
of time: The organisms carry out these reactions by means of enzymes and some
of the basic reactions that can occur are as follows:
Hydrolysis The addition of H20 to the molecule. The next step in
molecular breakdown is often thus facilitated so that
molecules are split at the hydrolysed link.
Dehydration The removal of H20 from one or more molecules.
Oxidation The removal of hydrogen or the addition of oxygen to
the molecule; also a process which involves an increase
in the number of positive charges on an atom, or a
decrease in a number of negative charges.
Reduction The removal of oxygen or the addition of hydrogen; the
reverse reactions to those of oxidation.
Decarboxylation The removal of CO2.
Dea,nination The removal of —NI-I2.
Preservatives 677
Preservatives 679
.2 CC
. .2
C)
aCC CM
CC
a -4
0 N
C
v
o c
a - CM —
CM
CM
•0 •0
0
I-.
0
I-
C) U
C) >
CM
CC CC
CO
CM C)
C) V
C) a 0
C C) >
a 4)
. 0 2 2
CC
CC
I- I-. 4) .0
00 TS
04)
(C
C co
'-.• C- I-
C) CC
•0 >
4) C) .a C)
CM
a .0 Cd
o
C)
0 0 OOo 0
0 . 0 a
tI 0 0. . 0 z
C a
C)
a a
a cq
411 CC CC
CC . 1 0
CM CM
C) C) o
C .0CC
C) C) 0. .
a.a C) CC
CM
0.
C)
CM C .
0.
CM
iJ
4.)
0 CM
2
;
a
0 •CMC 0
a
CM NF Z
0
CM
0 C 1 OCCcJCC
c) I-
C C
0
CM
a' en en
—4
0 C
0
U
CM
'I
-C) •0 -C) 0
C) 41 C) C) C) C) C)
CM CM
0
CM CM
a a
a
a
a
a •0 a a
I-0
C)
E C)
CM •0
CM
a02 CM
CM
CC C) C)
•0 a 4.) (C
C)' QC)
4)4)
- . i... .-
- C) O
C) CO
0.>
CrC-fl
41 0.> a 0CM > CM
0CC
en U C) 0 C)
cC C- O .0 CC CO C) CO C)
'0
I-
680 R irry's Cosr zeticology
•0
0-
V 0 N
CM N - N N
L. CC a' . . 0 0'
I.. -
>'.
0 CC
CC
-C pq
I,
C)
C)
CC .0
CM • 0. 'a
0 0 0..
CC 0.CC
0 . 0
,..CC .9 (C
Ia
CCC)
CC
.L1
0
C) CC
0.
' .
IaC,C 0.
C) CM
0 >
C) 0
•0
Ca C) O 0. C)
.0 CC
II.
0
CM
o 0 11 .0 CC
C-.
V 00 C) .0
-. ci
u
V o C)
0•
.0CM 0.
CC
u >C C)
0 CC0
V 0.-C. v
CC
CO CC CC
0
.0
0 •
0' C) 0 '-
C) c
CC
CC,.. 0
cj0 O.CC 00
0.D . 00 E
_;C C) •..0
000 . S
U 0
m• CC. 0 .0 .0 U V
ct
CC
*
CL 0. 0
0 C) 0 0. CC S
C 0 0..
•0
I
I I
CD
c In en
—I
V
0 I
O I
•0
I C). IC) C) C)
CM
•
'a 0 0 0
0 CM
(C 'a C)
C)( •0 C-. C- i'a I"
00 p CC C" CC1
-4 CM 0
CC I0I
CM C..
.0 CO...
V i0 I
C..VVC)EE a.a 0, '
V CM I
0.0. CC 0.>0
•0
0 CCCC o 9
I I-' I
CC
0
I- 00U N0 C
a' a's
— C4fl I N
-C I
—C I
Preservatives 681
Trichothecium
Verticillium
Origins of Contamination
Raw Materials
If the raw materials used in manufacturing of cosmetics are heavily contamin-
ated then it is almost inevitable that the finished product will be similarly
contaminated, and any preservative system present will be unnecessarily
can be avoided by careful monitoring of raw materials.
The water used in product manufacture is possibly the most frequent source of
contamination, often supporting large numbers of micro-organisms. Mains
water contains low numbers of organisms, usually less than 300 ml'. Softened,
distilled or, in particular, deiouized water is capable of supporting the growth
of some bacteria, and numbers can rise to 106_107 ml when water is
stored . 2325 Storage tanks are thus frequently responsible; Baker" describes
several cases of serious product deterioration which were traced to heavy
contamination in the beds of ion exchangers or water tanks. More recently
Duke 3 describes the contamination of a hair rinse v ith waterborne Gram-
negative organisms which were using cetrimide in the product as a nutrient.
Correlations between shampoo spoilage and waterborne organisms have been
discussed .21
Fats, waxes and refined oils contain relatively few organisms whereas natural
materials such as gums and herbs are frequently very heavily contaminated by a
variety of fungi, yeasts and bacteria. The natural gums, tragacanth, karaya and
682 Harry's Cosmeticology
acacia, are always very heavily contaminated; the synthetic gums are often
almost sterile. Other naturally occurring materials such as talc, kaolin, chalk and
rice starch frequently carry large numbers of bacteria, particularly those that are
capable of forming spores. Duke reports contamination of a face preparation by
a pigment contaminated with bacteria. The contaminants were killed by addition
of a preservative.
The containers of raw materials—drums, sacks, cartons, etc.—may also be a
source of contamination prior to manufacture.
Environment
A further possible source of contamination is the air, which contains mainly
mould and bacterial spores and skin cocci. Environmental control is facilitated
by covering of containers and reduction of air-currents over a product.. 23 Routine
monitoring of the air and selected surface sites within the production area is
advised, so that deviations from normal standards of cleanliness are immediately
detected)
A survey of the types of organisms found in manufacturing environments and
how these varied over a nine-month period has been conducted by .Bruch.27
Equipment
During manufacture the product can easily become contaminated by organisms
which accumulate in the plant as a result of faulty or inadequate cleansing.
Pieces of equipment with inaccessible joints, pipes and pumps are often difficult
to clean properly and washing with detergent solutions may only result in
dilution of the product to form stagnant foci where bacteria and fungi thrive.
Plant should be designed to facilitate easy cleaning and disinfection, and
inaccessible grooves and dead-ends in all items which come into contact with the
product should be avoided wherever possible (see Chapter 43).
An example of product deterioration resulting from plant contamination is
given by Duke. 3 A broken sealing joint in a bulk tank used for the manufacture
of a shampoo allowed shampoo and contaminated cleansing water to enter a
false bottom of the tank. As the preservative had been diluted, the contaminants
survived and were washed into each new batch of shampoo. Using the shampoo
base as nutrient, the bacteria then contaminated the whole product.
The need for good housekeeping procedures cannot be stressed too strongly.
Of paramount importance is the education of plant operatives to appreciate the
necessity of properly performed plant cleaning operations.
Advice on methods of plant cleaning and disinfection are given by Davis 29
who states that 150-200 ppm of hypochlorite will sanitize clean metal and glass
In two minutes and sterilize in 10 minutes. The corrosive nature.of this treatment
is a disadvantage and formaldehyde is frequently used in preference. Hot water,23
or preferably live steam, has been recommended as the best disinfecting agent.
Detergent sanitizers containing quaternary ammonium compounds or deter-
gent–iodine mixtures are also useful, but it is essential to remove product residue
from the plant and to rinse thoroughly with hot water before utilizing any of the
above disinfection procedures since many are inactivated by residues of organic
Preservatives 683
material. Air locks in the plant may also prevent certain parts coming into
contact with the cleaning and sterilizing fluids and should be avoided.
Microbiological tests should be made on equipment and filling lines before
and after sanitization to ascertain the effectiveness of cleaning procedures.
RODAC plates and swabbing techniques may be used for this purpose together
with sampling of the final rinse fluid after sanitization to determine the
cleanliness of internal surfaces.1
Packaging Materials
Since most products are exposed to further contamination during filling into
containers, cleaning and disinfection of filling equipment is also important (see
Chapter 43). Containers and closures should be dust-free and microbiologicall.y
clean. This can be achieved by a filtered air blast which is probably more
effective than the use of detergent and water. 13 Caps and liners are notorious for
an abundant fungal flora and much of the growth on the surface of cosmetic
creams is directly attributable to cap linings. Quite frequently creams containing
preservatives, which are otherwise adequately preserved, break down following
the introduction of large numbers of fungal spores from a cap lining. These
organisms germinate initially in the microfilm of water on the lining and spread
gradually to areas of slight separation around the edge of the cream. Many toilet
preparations, in addition to having a long shelf-life, may be in use by the
consumer for many months, so that spoilage does not become apparent until the
contents are more than half used. Products packed in wide mouthed jars, and
flexible bottles which draw air back into them, are more liable to contamination
than those packed in collapsible tubes and bottles with small orifices.
As plastic materials are not subject to biodegradation, use of such materials
instead of cellulosic materials should be associated with a reduction in microbial
spoilage. Unlike paper, card and cork, plastics are microbiologically pure but as
they are porous to oxygen and carbon dioxide and encourage water condensa-
tion, spoilage by mould spores will be facilitated .7
Personnel
Probably the greatest microbial hazard to the product during manufacture or
packaging is from the operators .23,29 Operators should be properly instructed to
appreciate that they are a potential source of contamination and trained to
maintain high standards of personal hygiene and cleanliness. The wearing of
protective overgarments is advisable.23
Water Content
Because micro-organisms depend on water for the synthesis of cellular compo-
nents, the physical and chemical characteristics of the water phase in an
emulsion, for example, are dominant factors in determining whether growth
684 Harry's Cosmeticology
pH Value
The pH value of a product will affect the degree of ionization of utilizable
materials, influence the electrical charge at the bacterial and fungal cell walls,
determine enzyme production and activity and hence regulate the availability of
nutrients and the ease with which they are assimilated by the microbial cell.
However, since growth tolerance limits for pH differ widely for various
micro-organisms, pH itself should never be regarded as a factor likely to
contribute to self-sterilization of a product. Pseudomonas species, which are
extremely common contaminants of toilet preparations, can exist over as wide a
pH range as 30-110 and although many fungi grow most prolifically at acid pH
they have been known to survive on vanishing creams at pH 9.
Osmotic Pressure
The living semi-permeable membranes which surround the bodies of all micro-
organisms can be ruptured by changes in osmotic pressure and this can lead to..
membrane shrinkage and dehydration of the organism. For this reason osmotic
pressure can have a limiting effect on growth. Concentrations of between 40 and
50 per cent of glycerin and sorbitol are inhibitory to micro-organisms by reason
of the osmotic pressure, and high concentrations of electrolytes can exert a
similar limiting action. Thus very concentrated products are likely to be
self-preserving but those which may be diluted and left standing before use can
deteriorate if left for periods of several days. Shampoos, for example, which are
frequently sold to professional hairdressers as concentrates for dilution before
use, are often sisceptible to bacterial degradation as the concentrates may be
diluted with contaminated water and left-uncovered for long periods during use.
Surface Tension and Oxygen Tension -
Quite apart from the nutritive value of low concentrations of some surfactant
molecules, surface tension is itself a factor influencing growth. Many Gram-
negative bacteria, and the coliforms in particular, grow well in environments
abounding in surfactants, while most Gram-positive organisms do not grow well
at surface tension levels much below 50 dyn/cm (0 . 05 N/rn). Gram-negative
organisms, particularly the 'seudomonads, flourish in shampoos and are also
common contaminants of the aqueous phases of emulsions. Cationic suyfactants
are toxic to many organisms, anionics to a few, and nonionics to hardly any; thus
su?lte tension per se will not be a gross limiting factor but will have an effect in
association with the presence or absence of toxic groups in the surfactant
molecules. Most organisms, bacteria and fungi which contribute to product
spoilage are aerobic and depend on the availability of oxygen for their
metabolism. The microclimate in most products, with perhaps the exception of
those in pressurized packs, will almost invariably provide sufficient oxygen for
the growth of micro-organisms provided that all other factors are favourable.
Antimicrobial Spectrum of the Preservative System
In a cosmetic, as in the natural environment, there is competition between
organisms for nutrients so that one organism will survive and grow more
successfully than others. Thus where a preservative is active against a limited
spectrum of organisms, the product may be left open to spoilage by less
susceptible organisms.36
686 Harry's Cosmeticology
Temperature
The susceptibility to microbial attack will vary with the temperature of storage,
so that a cosmetic kept at room temperature will be liable to spoilage by different
organisms from those that flourish in a product kept in a hot environment
(for example one left in the sun or in a hot car). Bacteria generally prefer
.31
temperatures of 30°-37 o C and fungi and yeasts 20'-25'C
Preservative Requirements
The 'ideal' preservative which is both safe and effective in all kinds of toilet
preparations has not yet been discovered and this means that the composition of
each new product must be studied in detail before selecting a suitable preserva-
tive. In order to avoid preservative failure a careful analysis must be made of the
factors in the product that are likely to favour the growth of micro-organisms,
the ingredients that are likely to be contaminated before use, and also those that
are likely to influence adversely the efficiency of' whatever preservative is
ultimately selected. The essential requirements of a preservative are:
(i) freedom from toxic, irritant or sensitizing effects at the concentrations used
on the skin, mucous membranes and, in the case of orally administered
products, on the gastro-intestinal system;
(ii) stability to heat and prolonged storage;
(iii) freedom from gross incompatibility with other ingredients in the formula
and with the packaging material, which could result in loss of antimicrobial
action.
Other requirements are that the preservative should be active at low concen-
tration; should retain its effectiveness over a wide range of pH; should be
effective against a wide range of micro-organisms; should be readily soluble at its
effective concentration; should have no odour or colour; should be non-volatile;
should retain its activity in the presence of metallic salts of aluminium, zinc and
iron; should be non-corrosive to collapsible metal tubes and non-injurious to
rubber.
Table 36.3 lists some of the preservatives used in cosmetics and toilet
preparations. Compcunds closely related to those listed in Table 36.3 have also
been used, and inspection of the variety of preservatives available shows that a
thorough understanding of the factors that can influence the efficiency in a
particular system is necessary before a selection can be made.
The properties of some individual preservatives have been discussed by
Table
Gucklhorn,37 Rosen and Berke, Croshaw38 and Cowen and Steiger. 39
36
36.4 indicates the advantages and disadvantages of some of the established
groups of preservatives. -
practice many preservatives are pH-dependent, the majority of them being more
active in the acidic than alkaline range. Some preservatives with a wide pH
profile have the disadvantage of being chemically highly reactive compounds
(for example formaldehyde and formaldehyde donors) which react with other
components of the formulation .39 pH may also have an effect on the microbial
cell surface" and may affect the partitioning of an antimicrobial agent between
the cell and the product.4'
For many preservatives the most pronounced effect of pH on activity is on the
antimicrobial agent itself. Many weak acids are used s preservatives, their
activity depending upon the amount of undissociated acid, which in turn
depends on the dissociation constant and pH of the system (see Tables 36.5 and
36.6). It has been suggested that anions of acids may be inactive as a result of
—45
688 Harry's Cosmeticology
an
C-) an
'I) 0)
C'S a)
a)
-a) U
U .0
I-
-
0 C.)
a) E
0 C) .0 E atU
C)
I.- 0 00) C
cc
an 0 V a)
I-.
C) o- a), cctaOCn)
-c a) .-
a-.0) It a), >0
C) E o a)
0) 0 o 0 - U 0.2
0. a)
0
.0
"So
(11 8
a)
S.-.
Oat
Ca)
0)
an
C'S - .
an C) • N; . .0
Eano a) .-a)a) a)
S.-.
a- .-. a)
to
0 .0 0) .E 2.9 .0 )
•0
a) C. a) a) .0
I-
at .2
a) at
c-I bOa) U)
0 cc
C) Ur 4
C-) a) -
cc 8a • a)
cc a)- C)
.-C)
-E .g
:2: E o.
o 0 S..
an
cc b .- 0 C'S at '6.-° 0 o.0oCt
'S . .0 o
00 ,
C
0000
. .
C
an
a)
an an C
a)
ata-. •0
a)
0 U
at "S
•0 U
0C) an
a,)
>
0
U
a) o an..
"S
S.-
a)
U a)
U,
a)
0) C) a)
ala 0.
a) 00
a) tan tO)
.E a-
a)
Ct .E t
E
00
E E > 0)
C)
C); a) a) U
a-.
an o Coat
a)
C) C) 0. a)
C'S 0)
0. 00
co
. •6
'aI'
•- 0. at=
an
a) - 0) at
>
at
Cat
2 U
4: L)
.000 a),
C.)
C) C) 0
a) at C) "S -a)
Uo'S
CN
.00. 6 C
N 2 -C
0)
• 0. a) > C)
000 .0
U 15:1
.; 0 0)0
U,
a), •0 .0 U
'C a) C-) .
-o -0
:1 at 0)-a) C
'a E
.0
"S.
at a)
0)
O ct a), C, .0
a)..
4: 0
Preservatives
689
Table 36.5 Dissociation Constants of Acids used as Preservatives
Dissociation
Preservative constant
Sulphurous acid '2S03 - 1 . 70 )< 2
o-Chlorobenzojc acid o-Cl . C6H4COOH 1•20 x
Salicylic acid o-HO C6H4COOH 106 x iO
Formic acid H.COOI-1 180 x io -
p-Chlorbcnzoic acid p-Cl CH4COOH 1•05 x
Benzoic acid C6H5COOi-t 63() x iO
p-Hydroxybenzoic acid p-HO C6}-I4COOH 3O0 x io-
Sorbic acid CH3CH: CHCH : CHCOOH 1•73 x l0
Propionic acid C2H5COOH 1•40 x 10-5
Dehydroacetic acid OCOCH(COCH 3 )COCH : C(CH3) 530 x 10-6
repulsion from the negatively charged microbial cell wall. Benzoic acid is an
excellent preservative in its undissociated form but is strongly pH-dependent, so
that at pH 6 approximately 60 times as much benzoic acid is required as at pH 3.
Similarly, sorbic acid is present mainly in the undissociated (active) form at pH
4, but only 6 per cent of this form is present at pH 6. A chlorine addition product
of sorbic acid 4, is claimed to be more effective than sorbic acid itself and less
affected by high pH.
Dehydroacetic acid enolizes to give a weak acid and has been studied
extensively as a preservative,' particularly for foodstuffs .43-45 It has a very low
dissociation constant which indicates that it retains activity at higher pH values
than most other organic acids and for this reason it commands wider usage. At
pH 6, 16 per cent of the dehydroacetic acid is in its undissociated form, which is
more than benzoic acid at pH 5. The anion of dehydroacetic acid is said to be
weakly antimicrobial so that at pH 7 this acid will retain some activity .36
Phenols, which include the Parabens, behave as weak acids and consequently
are less dramatically affected by pH than stronger acids. For example, methyl
Paraben at pH 8.5 is approximately 50 per cent undissociated.
The relationships between pH and effectiveness of a range of preservatives
have been studied by a number of workers 4648 Simon 48 showed that the
biological activity of weak acids is influenced in a regular manner by changes in
690 Harry's Cosmeticology
Concentration of Preservative
There cannot be any hard and fast rules about the optimum concentrations at
which various preservatives should he used. This will be obvious from the
foregoing paragraphs which mention the multitude of factors that contribute to
the growth of organisms in products and the effectiveness of preservatives in
those environments. Some products, by virtue of the concentration of materials
in their aqueous phase, are virtually self-preserving without any addition of
preservatives being necessary, whereas others may provide a nutritious environ-
irient for the growth of micro-organisms and thus require a fairly high concentra-
tion of a powerful preservative.
Effective concentrations of preservatives range from as little as 0001 per cent
in the case of organic mercurial compounds to 05 per cent or even I per cent of
such materials as the weak acids, depending on the pH of the product.
The availability of the preservative to the micro-organisms it is required to
inhibit is probably more important than the overall concentration itself. 'Availa-
bility' in this context can be defined in accordance with the mechanism of action
of the particular preservative, and may depend on the permeability across the
cell wall (if this is the mechanism), the flux (if diffusion rate is important) or
degree of adsorption (if the preservative acts by coating the surface of the
organism). The availability of the preservative is also influenced by the
distribution, or partition, of the preservative between phases of the product.
Partitioning properties of preservatives are discussed later in the chapter.
There are certain advantages in using preservatives in combination rather than
singly, which are as follows: (a) a broadening of the antimicrobial spectrum of
Preservatives 691
activity; (bT the use of a lower concentration of each of the preservatives, thus
avoiding problems of toxicity or insolubility; (c) a reduced probability of survival
of an organism resistant to one of the preservatives, provided that the other
preservative(s) in the system act by a different mechanism; (d) the antimicrobial
activity of the combination may be greater than the additive effects of the
individual preservatives.
Frequently the esters of p-hydroxybenzoate arc used in combination, the
methyl ester in the aqueous phase of an emulsion and the propyi ester in the oil
phase. Addition of the preservative to the oil phase is not so much to prevent the
multiplication of organisms in this phase, as this seldom occurs, but to prevent
diffusion or partition of the methyl ester from the aqueous phase into the oil
phase; the presence of propyl ester in the oil phase will tend to stabilize the
distribution between the phases. Bean et al. 50 have examined the activity of
phenol against E.coli in oil—water dispersions and determined the distribution of
phenol between oil and water in liquid paraffin and arachis oil dispersions.
Determination of the extinction times of the organism in the systems showed
that the bactericidal activity was governed by the concentration of phenol in the
aqueous phase and the proportion of oil to water.
The Parabens have been used in combination with other preservatives, for
example phenoxyethanol. This combination (marketed as Phenonip) is claimed
to have a wide antimicrobial spectrum of activity, the activity being retained in
the presence of surfactants and proteins. 38 A synergistic effect has been reported
between Phenonip and hexachlorophene, cetylpyridium chloride, thioniersal or
dichlorophenc. 5 ' Imidazolidinyl urea has been found to act synergisticall y with
other preservatives including methyl and propyl Parabens, increasing the
antimicrobial capacity of the preservative system and the spectrum of antimicro-
bial activity.36
Synergism has also been found to occur with combinations of henzalkonium
chloride or chiorhexidine with some aromatic -alcohols. 52 Hugh0 53 has reported
synergism between p-chloro-m-cresol and bcnzalkonium chloride, m-cresol and
phenylmercuric acetate, and bcnzalkonium chloride and phenylmercuric ace-
tate.
Partition Coefficient
The preservation of formulations containing oil and water is complicated by the
ability of preservatives to distribute themselves between these two phases. Since
micro-organisms only grow in the aqueous phase it is important that the
preservative does not distribute itself in such a way as to leave an ineffective
Concentration in this phase. Ideally a preservative should have high water
solubility and low oil solubility, that is, have a low oil—water partition coefficient.
For simple systems where no emulgent is present the concentration of the
preservative in the aqueous phase (C) can be calculated from the following
equation:
Cw- C((+1)
- (K + 1)
where C is the total preservative concentration, 0 is the oil—water ratio and K. is
the oil—water partition coefficient. 54 The concentration of preservative in the
692 Harry's Cosmeticology
in certain preservatives and the basic oxygen in the ether group of the ethylene
oxide adducts. This mechanism cannot, however, account entirely for nonionic—
preservative interactions since a high degree of hydrogen bonding, and conse-
quently inactivation, would be expected when carboxymethyl cellulose and gum
tragacanth are present, but these materials do not inactivate preservatives to
anything like the same extent as the high-molecular-weight polyethylene glycol
ester, for example.
Coates and Richardson 72 have examined the activity of cetylpyridinium
chloride in aqueous solutions of polyethylene glycol and found that, except for
very high concentrations, activity is reduced by the presence of the glycol,
although not as much as predicted by binding data.. Since polyethylene glycol
does not form micelles, these workers suggest that the interaction may be due to
an attraction between the electron-deficient pyridinium ring of the antibacterial
agent and the electron-rich polyether linkages of the glycol.
Preservatives that appear to be much less affected by the presence of nonionic
surfactants are formaldehyde, sorhic acid, benzoic acid and dehydroacetic acid .57
Nonionic surfactants form micelles in aqueous solutions at very low concentra-
tions and for this reason, when used either as emulsifiers-or solubilizers, will
always be present at concentrations well above their CIVIC. In order to be
effective, preservatives must be in solution and 'available' in the aqueous phase
of a product and the hydrophile—lipophile characteristics of the preservative will
influence its relationship with the nonionic. The more lipophilic preservatives
appear to be bound to a greater extent than the more water-soluble compounds,
and Patel and Kostenbauder 73 have studied the effect of Tween 80 on methyl
p-hydroxybcnzoate and propyl p-hydroxyberizoate. The propyl ester was found
to have a far greater affinity for Tween 80 than the methyl ester. At 5 per cent
Tween 80, 22 per cent of the methyl p-hydroxybenzoate existed as free preser-
vative; under equivalent conditions, only 45 per cent of the propyl p-hydroxy-
benzoate existed in a free state. The interaction between the nonionic surfactant
cctomacrogol and benzoic acid, p-hydroxybenzoic acid, methylparabens,
propylparabens and chloroxylenol has been investigated and 6partition and
binding phenomena discussed by Kazmi and Mitchell. 65 Konning6 de scribes the
interaction between phenol or chlorocresol and polysorbate 80 in arachis
oil—water systems, and discusses the effect of altering the surfactant concentra-
tion and the proportion of oil on the concentration of preservative in the aqueous
phase.
Baley et al. 67 have investigated the bactericidal properties of some quaternary
ammonium compounds in dispersed systems. The concentration of quaternary in
the aqueous phase was varied by using different hydrocarbons, different
concentrations of hydrocarbons and different surface-active alcohols, and it was
demonstrated that the bactericidal activity corresponded to the concentration of
free preservative in the aqueous phase.
Attempts have been made to describe the systems mathematically in ordci
that the quantity of preservative required to produce effective preservation in a
surfactant solution or emulsified system may be calculated. Kazmi and Mitchell63
have derived separate equations for surfactant solutions and for emulsified
sy stems which were found to correlate with experimentally determined values.
In another paper 64 they developed the theory of capacity. In a sotubilized system
Preservatives 695
a,
oN
C N
0
a, C
.0 a,
N >.
•0
>
Macromolecule (%
Figure 36.1 Ratio of total to free preservative concentration in the presence of different
concentrations of macromolecules (R value of 1 . 0 corresponds to no measurable
binding61)
greater adsorption will occur than if it is added after the particle surfaces have
become coated with surfactant.
The activity of a preservative may be reduced by interaction with, or loss
through, the container or closure. Interaction of preservatives with rubber has
been well documented . 44 ° There is increasing interest in the interactions
between preservatives and plastics. Phenolic and quaternary ammonium com-
pounds have been reputed to react with polyurethane." Parabens, benzoic,
sorhic and salicylic acids are taken up by nylon, polyvinyl chloride and
polyethylene .78
Selection of a Preservative
Although it is not possible to lay down a set of 'rules' to which a formulator
should adhere, weeks of work can often bcsaved by considering on a theoretical
basis the factors that are likely to influence the preservation of a new product.
697
Preservatives
This approach, coupled with simple laboratory tests on various combinations of
the formula components, while not a substitute for thorough microbiological
testing of the finished formula, can save time and frustration.
The complexity of modern formulae often means that there is a variety of
materials present, some of which will act in favour of good keeping qualities
while others will act against whatever preservative is chosen; the relative
hostility of micro-Organisms or the nutritive value of the formula itself is also of
importance.
Obvious steps to follow before selecting a preservative are listed below.
698 Harry 's Cosmeticology
'I
—v
Gd =
II I
U)
—
3rCN NrC
U-'
C)
'-I
N
N
., .'
C-
C
C
I—
C
U
0
C-)
C C:
o -' N r' N N C
c —rI0 'C
H C-)
C
-c
'-. C C CC
0r
el C)
C:
H
CU)C
X,c
-C--
Cd
N >,C >
'-c L)Q
I-
Gd V
z -E U)
Cd = - C .=
I-
699
Preservatives
materials arc:
Bronopol (2bromo2nitrO-1,3Pr0Pane db0 O
which, according to Croshaw et
Pseudotnonas species, is only slightly
al.,' 9 is active at low concentrations against
reduced in activity by nonionics and has low toxicity.
Chiorhexidine ( b is (pchloroph CflYldiguanido)hene),
is a wide-spectrum anti-
microbial agent with a good record of safety.
DehydroacetiC acid, which is suitable for formulae of low pH, is relatively
unaffected by the presence of high levels of nonionic emulsifiers and appears to
he safe for use on the skin.
Inzidazolidinyl urea, which is not pH-dependent has high water solubility, is
non-toxic, non-irritating and non-sensitizing. It is active against Gram-positive
and Grain-negative bacteria, but selectively active against yeasts and moulds. It
retains its activity in the presence of many cosmetic ingredients, including
surfactants, and frequently acts synergistically with other preservatives, for
example Parabens.3
Mixtures of preservatives that are effective against different micro-organisms
are also often useful, For example -p-phenoxy-ethyl alcohol, which is highly
active against Grain-negative bacteria and fungi, can he used with a quaternary
ammonium compound such as benzalkonium chloride which acts against Gram-
positive bacteria at very high dilutions. In addition there are advantages in using
combinations which are no only active against a wide range of organisms but
also act synergistically, for example imidazolidinyl urea and Parabens.
Safety Aspects
Preservatives are commonly expensive ingredients and it is always advisable to
use the lowest effective concentration. The cost, however, is secondary to the
more important question of safety to the consumer.
If tests to measure the effectiveness of certain preservatives show that several
times the usual concentration is required to achieve the desired antimicrobial
action (because of increased partitioning into the non-aqueous phase, physico-
chemical binding or factors influencing dissociation), it is wise to consider the
toxicity of the preservative at the higher concentration before proceeding.
Although a preservative may he partially bound in a product and the remaining
fraction in the aqueous phase may represent no more than is safely used in other
formulae, the ratio of bound to free preservative is unlikely to remain tin-
changed when the product is actually in use. Thus from a toxicity point of view
the total amount is important rather than only that fraction which is acting as a
preservative in the particular vehicle. Application of the product to the skin, for
example, will disturb the original preservative equilibrium between the oarious
phases of the product and will almost certainly result in liberation of the
preservative previously bound. Evaporation of water will increase the concen-
tration of the preservative available to the skin and may result in primary
irritation or, in some cases, sensitization.
However, rather than a sharp dividing line between a toxic and a non-
toxic concentration of preservative there is, instead, a reasonably continuous
700 Harry's Cosmeticology
higher concentrations can cause erythema and drying of the skin. Cases of
sensitization to cetrimide at concentrations of about 1 per cent have been re-
ported by Cruickshank and Squire. 86 Their substantivity to the human skin has
caused concern about plant safety. 87
Formaldehyde is well known to be a skin irritant, and for this reason, and for
reasons of volatility and odour, it has not been used extensively as a preserva-
tive. In some countries such as Japan and Sweden its use is banned :36 Slow
formaldehyde-release agents appear to possess less sensitizing potential.39
The toxic thresholds of preservatives will depend not only upon the concentra-
tions at which they are used but also upon the vehicle. A certain concentration of
a particular preservative may be quite harmless in one system while the same
level might evoke adverse skin responses in another because of the presence of
substances which increase its penetration through the skin.
Fungi Yeasts
Because all the factors governing preservative effectiveness are not at present
completely understood, it is almost impossible to predict with certainty whether
a particular preservative in a particular system will be efficient. Consequently
during the fairly early stages of formulation some microbiological testing must be
undertaken to check the compatibility of the preservative in the new system.
Inoculation Tests
Methods by which known numbers of bacteria or fungi are introduced into the
product and samples taken at intervals to estimate survival are by far the most
reliable. Various test procedures have been proposed 6,118.81 and have been
reviewed by Cowen and Steiger. 9° In general, ascertaining the resistance of a
product to bacterial contamination involves inoculating a suitable size sample
(for example 10 g) of the product with the test organism to give a final
concentration of iO— 107 organisms g . The number of survivors in the sample
is determined at intervals after storage at room temperature. The standard
which must be met in order that a product may be considered effectively
preserved varies depending on its intended use. Differences in opinion regarding
interpretation of results are reflected by the different standards set by the USP
test, the Society of Cosmetic Chemists test and the Toilet Goods Association
test.
Longer-terms tests, in which smaller numbers of micro-organisms are used
and the samples are observed for changes in their physical characteristics over
several months, may be more meaningful. Products inoculated with spore-
bearing organisms can only be observed for physical changes, as microbiological
sampling is unrealistic since spores which might remain dormant in the product
will germinate when transferred to nutrient culture medium. In samples
inoculated with vegetative organisms, a gradual diminution in numbers can he
traced over a period of time if the preservative is effective, but with fungi and
spore-bearing organisms one can only wait for theappearance of visible spoilage
and this sometimes takes several months to occur.
Inoculation tests have been known to produce misleading results when
organisms, artificially introduced into a product, have died off in a short time
and the product has been considered to he adequately preserved. Then, later,
perhaps during manufacture at the factory or on storage, the product has shown
the effects of contamination. One reason for this is that the wrong types of
organism were used for the test, and time and the right conditions were not
provided for adaptation of the organisms to their environment. Test organisms
should be chosen to represent the types of organism that are known to be
frequent product contaminants, for example Pseudornonas species, and
those with which the product is likely to come into contact, for example
Staphylococcus species. In addition, organisms isolated from the manufacturing
environment, raw materials and, where possible, from contaminated products
(preferably of the same or similar formula) should be used. In the testing of
shampoos, for example, tap water provides a suitable source of test organisms.8
The maintenance of test organisms is crucial since their resistance is influenced
markedly by the medium on which they are grown. To ensure suitable
resistance, test organisms may be grown in a medium containing the preservative
Preservatives 703
or product in low concentration. 91 Product contaminants may be kept in
unpreserved or inadequately preserved preparations. 88,92
Mixed cultures may be used initially to reduce the amount of testing required
to assess the adequacy of the preservative system, while pure culture chbllenge
may be employed to give more detailed information about preservative ade-
quacy against specific organisms. 93
Most tests run for a minimum of 28 days, the product being sampled for viable
organisms at various intervals, depending on the probable frequency of usage
during this period. Slow adaptation of micro-organisms to their environment
makes it essential to test for long enough to determine whether inoculated
bacteria and fungi will grow after a dormant period, and in some instances tests
lasting as long as six months may be too short.
Tests should also be performed on products that have been stored for specified
time intervals at temperatures and humidities which the product is likely to meet
during use, in order to ensure that adequate preservative activity is retained
throughout the shelf-life. 92 A concurrent chemical assay of the preservative gives
additional valuable information.
Products may be tested using a single inoculation or using an inoculation–
sampling cycle. 8 The latter method, in which the sample is subjected to more
than one challenge, has been advocated by several workers" since it is
considered to be more representative of in-use conditions and has the advantage
of indicating at what point the preservative system will fail. A criticism of the
capacity test is that it may lead to excessive preservation, with consequential use
of dermatologically unsafe preservative, concentrations. The balance between
over-preserving and under-preserving a product will depend on the number of
challenges, which can only be chosen after a period of experimentation.90
In all cases the challenge should take place with the product in the container in
which it will be used by the consumer. 92
Preservatives
A proposed amendment to the EEC cosmetics directive 95 contains a list of
preservatives for use in cosmetics and toiletries. This list is divided into twb
parts. The first part includes eleven preservatives that are considered acceptable
for use in cosmetics, while the second part is a provisional list' of 56
preservatives which may be considered acceptable subject to provision of further
information regarding their safety over the period up to 1982. At the time of
writing the amendment has not been incorporated into the directive.
REFERENCES
Developments in Industrial
1. Rodgers, J. A., Berka, S. Y. and Artest, E. G.,
New York, Plenum Press, 1974, p. 217.
Microbiology, Vol. 15,
2. Dunnigan, A.P., Drug. Cosme!. Ind., 1968. 102(6), 43.
Sxxxv.
3. Duke, A. M, J. app!. Bact., 1978, 44,
4. Allwood, M. C., J. app!. Bad., 1978. 44, Svii. New York,
5. Halleck, F. Ii., Developments in Industrial Microbiology, Vol. 12,
Plenum Press, 1971, p. 155. J. Soc. cosmet.
6. Van Abbé, N. J., Dixon, 11., Hughes, 0. and Woodroffe, R. C. S.,
Chem., 1970, 21, 719. J. Soc. cosmet. Chem., 1972, 23, 721,
7. Smart. R. and Spooner, D. F., J. Soc. cosmef. Chem.,
8. Flawn, P. C., Malcolm, S. A. and Woodroffe, R. C. S.,
1973, 24, 229.
9. Breach, G. D., 1. Soc. cosmet. Chem., 1975, 26, 315. Developments in Indus-
10. Sokolski, W. T., Chidester, C. G. and Honeywell, G.1962, E.,
p. 179.
trial Microbiology, Vol. 3, New York, Plenum Press,
10, 133..
11. Baker, J. 11., J. Soc. cosmet. Chem., 1959,
TGA Cosine(. J., 1969, 1, 34.
12. Wolven, A. and Levenstein, 1., TGA Cosmet. J., 1970, 2, 39.
13. Dunngan, A. P. and Evans, J. R., Am. J. Ophf hal.,
14. Wilson, L. A., Kuehne, J. W., Hall, S. W. and Ahearn, D. G.,
1971, 71, 1298.
Am. Cos,net. Perfutn., 1972, 87, 63.
15. Wolven, A. and Levenstein, 1., Can. J. pharm. Sci., 1973, 8, 19.
16. Myers, G. E. and Pasutto, F. M.,
17. Abeam, D. G., Wilson, L. A., Julian, A. 1., Reinhardt, D. J. and Ajehlo, C.,
New York, Plenum Press, 1974,
Developments in industrial Microbiology, Vol. 15,
p. 211. J. Soc. cosmet.
18. Jarvis, B., Reynolds, A. J., Rhodes, A. C. and Armstrong, M.,
Chem, 1974, 25, 563.
19. Baird, R. M., J. Soc. cosmet. Chem., 1977, 28, 17.
Developments in industrial Microbiology, Vol.
20. Ahearn, D. G. and Wilson, L. A.,
17, New York, Plenum Press. 1976, p. 23.
Preservatives 705
21. Morse,L. J., Williams, H. L., Green, F. P., Eldridge, E. E. a,pd Rotta, J. R., New
Engl. J. Med., 1967, 277, 472.
22. Morse, L. J. and Schonbeck, L. E., New Engl. I. Med., 1968, 278, 376.
23. Sykes, G., Microbiological control during manufacture, Society of Cosmetic Chem-
ists of Great Britain Symposium, Microbiological Safely of Toiletries and Cosmetics,
February 1976.
24: Goldman, C. L., Drug Cosmet. Ind., 1975, 117(7), 40.
25. Favero, M. S., Carson, L. A., Bond, W. W.-and Peterson, N. J., Science, 1971, 173,
826.
26. Malcolm, S. A. and Woodroffe, R. C. S., J. Soc. cosme:. Chem., 1975, 26, 277.
27. Bruch, C. W.,• Drug Cosmet. Ind., 1972, 110(6), 32.
28. Davis, J. G., J. Pharm. Pharmacol., 1960, 12, Suppl.29T.
29, Anon, Drug Cosmet. Ind., 1968, 103, (12), 53.
30. De Gray, R. J. and Killian, L. N., Developments in Industrial Microbiology, Vol. 3,
New York, Plenum .Press, 1962, p. 296.
31. Beveridge, E. G., Microbial Aspects of the Deterioration of Materials, London,
Academic Press, 1975, p. 213.
32. Bennett, E. C., Developments in Industrial Microbiology, Vol 3, New York, Plenum
Press, 1962, p. 273.
33. Barr, M. and Tice, L. F., J. Am. pharm. Assoc. sci. Ed, 1957, 46, 480.
34. Sawyer, C. N. and Ryckman, D. W., I. Am. Water Wk Assoc., 1957, 49, 480.
35. Yu-chih Hsi, Nature, 1965, 207, 385.
36. Rosen, W. E. and Berke, P. A., I. Soc. cosniet. Chen.., 1973, 24, 663.
37. Guckihorn, 1. R., Manuf. Chem. Aerosol News, 1969-171, 40-42 (passim).
38. Croshaw, B., J. Soc. cosmet. Chem., 1977, 28, 1.
39. Cowen, R. A. and Steiger, B., Cosmet. Toilet, ies, 1977, 92, iS.
40. Russell, A. D., Microbios, 1974, 10, 151.
41. Bean, H. S., I. Soc. cosmet. Chem., 1972, 23, 703.
42. British Patent 998 189, Farbwerke Hoechst iG, 5 October 1960.
43. Wolf, P. A. and Westveer, W. M., Arch. Biochem., 1950, 28, 201,
44. Von Schelhorn, M., Dtsch. Lebensm. Rundsch., 1952, 48, 15.
45. Von Schethorn, M., Fol. Techol., Champaign, 1953, 7, 97.
46. Bandelin F. J., J. Am. pharm. Assoc. sci. Ed., 1958, 47, 691.
47. Entrekin, D. N., J. pharm. Sci., 1961, 50, 743.
48. Simon, J., New Phytol., 1952, 51-52, 163.
49. Sykes, G., Disinfection and Sterilization, London, E. & F. N. Spon, 1965.
50. Bean, H. S., Richards, J. P. and Thomas, J., Boll. chins.-farm., 1962, 101, 339.
51. Boehm, E. E. and Maddox, D. N., Am. Perfum. Cosmet., 1970, 85, 31.
52. Richards, R. M. E. and McBride, R. J., I. pharm. Sci., 1973, 62, 2035.
53. Hugbo, P. G., Can. I. pharm. Sci., 1976, 11, 17.
54. Bean, H. S., Konning, G. H. and Malcolm, S. A., I. Pharm. Pharmac., 1969, 21
173S.
55. be Navarre, M. G., Chemistry and Manufacture of Cosmetics, Vol. 2, 2nd edn,
Princeton, NJ, Van Nostrand, 1962, p. 257.
56. Judis, J., I. pharm. Sc., 1962, 57, 261.
57. Wedderburn, D. L., I. Soc. cosmet. Chem., 1958, 9, 210.
58. Bean, H. S., and Berry, H., J. Pharntacol., 1951, 3, 639.
59. Bean, H. S. and Berry, H., J. Pharm Pharmacol., 1953, 5, 632.
60. Allwood, M. C., Microbios, 1973, 7, 209.
61. Kostenbauder, H. B., Am. Perfumer. Arom., 1960, 75(1), 28.
62. Kostenbauder, H. B., Developments in Indus,trial Microbiology, Vol. 3, New York,
Plenum Press, 1962, p. 286.
63. Kazmi, S. J. A. and Mitchell, A. 0., I. pharm. Sci., 1978, 67, 1260.
706 Harry's Cosmeticology
64. Kazmi, S. J. A. and Mitchell, A. 0.. J. pharm. Sc., 1978, 67, 1266.
65. Kazm,, S. J. A. and Mitchell, A. 0.,]. Pharm. Pharmacol., 1971, 23, 482.
66. Konning, G. H., Can. I. pharm. Sci., 1974, 9, 103.
67. Baley, G. J., Peck, G. E. and Banker, G. S., J. pharm. Sci., 1977, 66, 696.
68. Manowitz, M., Developments in Industrial Microbiology, Vol. 2, New York, Plenum
Press, 1962, p. 65. -
69. Wedderburn, D. L., Advances in Pharmaceutical Science, Vol. 1, London, Academic
Press, 1964, p. 196.
70, Schmolka, 1. R., 3. Soc. cosmet. Chem., 1973, 24, 577.
71. Tilbury, R. H., Specialities, 1965, 1(11), 3.
72. Coates, D. and Richardson, 0., J. app!. Bact., 1973, 36, 240.
73. Patel, N. K. and Kostenbauder, H. B.,]. Am. pharm. Assoc. sci. Ed., 1958, 47, 289.
74. Parker, M. S., J. app!. Bact., 1978, 44, Sxxix.
75. Poprzan, J. and De Navarre, M. G., I. Soc. cosmet. Chem., 1959, 10, 81.
76. McCarthy, T. L., I. Mond. Pharm. 1969, 4(12), 321.
77. Clarke, C. D. and Armstrong, N. A., Pharm. J., 1972, 211, 44.
78. Armstrong, N. A., Am. Cosmet. Perfum., 1972, 87, 45.
79. Crbshaw, B., Groves, M. J. and Lessel, B., I. Pharm. Pharmacol., 1964, 16, suppl.
127T.
80 Hjorth, N. and Trolle-Lassen, C., Trans. Rep. St. John's Hosp. Derm. Soc., Lond.,
1963, 49(10), 127,
81. Hjorth, N., Acta derm.-vener., Siockh., 1961, 41, (Suppl. 46), 97.
82. Sarkany, 1., Br. J. Dermato!., 1960, 72(10), 345,
83. Schorr, W. F., Am. Perfurn. Cosmet., 1970, 85, 39.
84. Fryklof, L. E., J. Pharm. Pharmacol., 1958, 10, 719.
85. Federal Register, 37 F.R. 12967, 30 June 1970.
86. Cruickshank, C. N. D. and Squire, J. R., fir. J. md. Med., 1949, 6, 164.
87. Smith, J. L., Cosmet. Toiletries, 1977, 92, 30.
88. Haileck, F. E., TGA Cosmet. 1., 1970,.2(l), 20.
89. United States Pharmacopeia XIX, Easton, Pa., Mack Publishing
1976,
Company, 1975.
27, 467.
90. Cowen, R. A. and Steiger, B., I. Soc. cosmet. Chem.,
91. Cowen, R. A., J. Soc. cosmet. Chem., 1974, 25, 307.
92. Moore, K. E., I. app!. Bacteriol., 1978, 44, Sxliii. CTFA
93. Cosmetic Toiletry and Fragrance Association Preservation Sub-Committee,
Cosmet. 1., 1973, 5(1), 2. Good Practice for the
94. Cosmetic, Toiletry and Perfumery Association Ltd, Code of
Toiletry and Cosmetic industry; Recommended Microbiological Limits and Guide-
lines to Microbiological Quality Control, 1975.
.95. Off. J. European Communities, 1979, No. C 165/52, 2 July.
Chapter Thirty-seven
Antioxidants
Introduction
The ability of atmospheric oxygen to act as an oxidizing agent for fats, fatty acids
and many other organic materials is of commercial importance. In some cases
the phenomenon may be advantageously employed, but in cosmetics the effects
of oxidation are normally deleterious and can lead to complete spoilage.
Although the literature on the chemical and physical changes involved in
oxidation is extensive and can be traced back to the eighteenth century' it is only
in the past thirty-five years that the mechanisms involved have been understood
with any degree of certainty.
Two of the problems associated with an understanding of the general
oxidation reactions have been the very wide spectrum of organic materials which
are subject to this type of decomposition and, secondly, the large number of
factors which can effect both the rate and course of the reactions. Amongst these
latter may be numbered the effects of humidity, oxygen concentration, tempera-
ture, UV irradiation and the presence or absence of anti- and pro-oxidants. In
the early studies of oxidative reactions many of the above environmental
conditions were not recognized as being important and were therefore. not
controlled, so that the results obtained are in many cases invalidated. Another
problem with the generalized oxidation theories which were developed from this
early work was that it was found difficult to apply the results from simple
reference substances such as methyl oleate or methyl linoleate to more complex
naturally occurring materials, such as sunflower seed oil or soya bean oil,
particularly at the high temperatures sometimes reached during the purification
or processing of these products. Other difficulties were encountered when the
general oxidation scheme was applied to organic materials other than fats and
their derivatives.
The problem with the cyclic peroxide oxidation theory was that only indirect
evidence for the existence of such compounds could be obtained. Most of the
evidence for their existence came from determinations of carbonyl, diene,
hydroxyl, iodine, peroxide and saponification values. These were considered
alongside determinations of molecular weight and the total oxygen uptake. The
problem was that the analytical results obtained were unreliable because the
oxidizing material was usually a non-fractionated natural product which would,
of necessity, show batch-to-batch variations in the proportion of individual
components and, as a consequence, exhibit variable oxidative behaviour. It was
also not realized that many of the analytical techniques used were not strictly
quantitative.
Alongside the ring peroxide theory existed the suggestion due to Fokin 7 that
the initial stage in the oxidation was the formation at the ethylene bond of an
epoxide of the type which can be isolated from many oxidizing peracids and
other oxidizing mixtures, but it is doubtful if these represent primary products in
the normal oxidation reaction.
The development of the modern theory of autoxidation may be said to start
.8
from the isolation of a cyclohexene peroxide by Stephens in 1928 On the basis
of theories then current he assumed that this was a saturated product with a
possible formula:
Cl-I2
-
CH2
I -
CH
CH2
CH/
IR02 + RH -- ROOH + R
R02 + k, non-propagating
Termination R02 + R - products
R + R' I
k
-.
where R, is the rate of formation1 of1/2chain carriers and k, k' are constants. It has
been shown that because (k4k6R ) is negligible in long-chain-length reactions
there would be complete correlation between the experimental and theoretical
710 Harry's Cosmezicology
equations:
—d[O,] = k3 k2(k4)''f 02]
R'12 [RHJ
dt (k4)"2' k3(k6)"[RH) + k 2 (k4)"[02] + (k4k6R,)"2
Un' 9 has applied stationary state conditions and produced the following
kinetic equations:
d[R]/dt = r - k,[R' ][O,] + k3[R0][RH]
d[R0]/dt = k2[ R ][0 2] - k,[Ro;][RH] - k4[R012
d [ROOH ]/dt = (r,/k4) 1/2 X (k,[RH])
At low oxygen level termination reactions involving R' become significant
and these kinetics would have to be modified to take account of this fact. Uri
therefore proposes a more general rate equation equivalent to that of Bolland17
and Bateman' for the formation of peroxide:
k2(k4)"2[O,]
d [ROOH ]/d: = (r,/k4)' 12 . k3[RH}
k 3 (k 6) ' /2[RH] + k2(k4)1/2[02]
He further suggests that provided k6 = k4 this may be simplified to:
k2f02J
d[ROOH]/dt = (ri/k4)"2 k1[RHJ
k a[ RH I + k2[02]
Harry's Cosmeticology
712
H H H
C C C. H
-H
/_ /_\
H
-
H
7 \/
C=C
H
CC
\ ,H H
/ \ CC
/
H HI-I / \
H
cis, cis cis, cis
trans, cis
02/
0-0
\/ "H
C\ H
HCC
H . CC'.
+H H CC
H. C=C -
/ \ / \
H H U
H
0-0
C H
/\ /
Temperature - H C=C
effect
H C=C
/ \
H
trans, trans
Figure 37.1 Oxidation of nonconjugated polyunsaturated systems (after Holman
the latter compound no peroxide occurred in the early stages. The reaction was
accompanied by the disappearance of double bond conjugation which was
directly related to the amount of oxygen absorbed, suggesting some form of
carbon-to-oxygen polymerization. Metal catalysts have also been shown to have
much smaller effects on the oxidation of conjugated systems, suggesting that the35
decomposition of peroxides did not play a significant role in their oxidation.
The relative rate of oxidation of the corresponding conjugated and nonconju-
gated systems is open to doubt and reports have appeared suggesting both
different and similar rates of oxidation, depending upon experimental
conditions. 34.36.37 The formation of polymer appears to cause steric hindrance,
thus reducing further oxidation, although the corresponding increase in viscosity
does not, as such, prevent oxidation even though it slows down the rate of
diffusion of oxygen into the system. The products of oxidation from these
systems are much more variable. Elcostearate oxidation appears to produce a
713
Antioxidants
The possibility that the perbenzoic acid was formed by a chain reaction which
could be influenced by light was discussed by B5ckstr6m 45 while later it was
shown that heavy metal catalysis could also occur .41 Waters and Whickham-
Jones47 found that the rate of oxidation was proportional to the square of the
benzaldehyde concentration and to the square root of dibenzoyl peroxide
concentration, and independent of oxygen concentration.
That other organic aldehydes oxidize through a similar free radical mechanism
has been shown by Cooper and Melville 48 using decanal. They showed that the
oxygen molecule reacted directly with the aldehyde and that with photoelectric
initiation the light was directly responsible for the formation of free radicals.
Oxidation of Ketones
The oxidation of ketones has been less studied than that of aldehydes and
appears to require a high temperature to produce decomposition. A tempera-
ture above 100°C was found by Sharp 49 to be required and the initially produced
hydroperoxide rapidly decomposed to give a mixture of acids and aldehydes.
Antioxidants
General Mechanism
If the general chain propagation reactions for oxidation are valid then it is
conceivable that the suppression of oxidation could occur either by suppressing
the formation of free radicals or by the introduction into the system of material
that would react with free radicals as they were formed, and so prevent a
build-up of reaction chains. The formation of free radicals cannot be wholly
prevented and therefore substances which behave as free radical acceptors-
antioxidants—are important.
Bolland and ten Have-50 studied the effect of hydroquinonc (AH) on the
oxidation of ethyl linolcate (RH). They suggested that this compound reacted
with the free radicals to give:
R+AH—..
I,
inactive products
R02 + AH
This led to the rate equation:
r = –d [0 21/di = rk3[RH]/k2[AH]
where r., is the rate of oxidation in the presence of the antioxidant and r, is the
rate of initiation of chains, and where the reaction R + AH is ignored as
being of little importance in the antioxidant reactions. However. Davies 5 ' has
assumed that the free radicals initially formed are R and thus the rate equation
in the presence of antioxidants would become:
r = –d[02]/dt = r(1 + k3[Rl-I]/k2[AH])
Thi0possibility that this equation is significant, at least in the early stages of
oxidation, has been cast in doubt by Bolland and ten Have who showed that,
when the reaction involved R0 2' radicals, the value (rjr2)[AH][RH] was a
constant with a value depending on the reaction coefficients k 4 /k 3 . k2 ; r is the
Antioxidants 715
The rate constant (k9) has been shown to be approximately the same for
different R02 radicals and is therefore of possible importance in the determina-
tion of antioxidant efficiency. 48 Except in the initial induction period and the
final stages of oxidation it was found that a plot of 11Ta versus time was a straight
line and that the rate of initiation was composed of elements due to temperature
and light intensity, so that:
re = k [RCHO ]10 21 + I
where I was the light intensity. -
A mechanism involving the intermediate formation of a complex between the
free radical R0 and the antioxidant has been proposed by Boozer and his
co-workers. 53 This would be followed by a subsequent rate-controlling reaction
with further R0 radicals:
rpid R0I
R0 + AH -. [R02 + AH] rotc onuoIIiag (k • products
Synergism
Synergism is said to occur when two or more antioxidants present in a system
show a greater overall effect than can be accounted for by a simple , addition of
their individual effectiveness. Although this phenomenon is well known, most of
the systems have been studied on an empirical basis. The phenomenon is
associated with two separate systems: (a) mixed free radical acceptors, and (b)
the metal cheLiting agents.
Mixed Free Radical Acceptors. It would appear that the effect of mixed free
radical acceptors is due to both steric factors and activation energy changes.
In a synergistic system involving a material such as ascorbic acid (BH) which
has a low stenic factor and hydroquinone (AH) in which steric factors would not
Antioxidants 717
be important, Un 59 has suggested that the following reactions will take place:
R0 2 + AH -. RO 2 H + A
A+BH —B+AH
The possible disappearance of A by reaction with oxygen is thus eliminated and
the effective antioxidant AH is regenerated. On its own BH would not produce
any significant antioxidant effect as the reaction:
R0 + BH—*RO 2 H + B
would be prevented by steric factors.
Metal Chelating Agents.. The normal effect of metal chelating agents is to bond
with pro-oxidant metallic ions and thus prevent their catalytic effect, on the
normal oxidation chain reaction. This reaction does not therefore prevent
normal oxidation taking place, but only slows down the formation of peroxide
while at the same time extending the induction period. Metallic pro-oxidants
that are already present as part of complex organic structures are not usually
affected by chelating agents. Stabilization has been achieved by reaction of the
metal with organic acids of the tartaric or citric acid type or with materials such
as ethylenediaminetetra-acetic. acid (EDTA).
Recent reviews of antioxidant–synergist systems have been published by
Prosperio 61 and Lozonczi.62
Typical antioxidants and synergistic systems.used in cosmetics are given in
Table 37.1.
peroxides are being formed faster than they are broken down. This condition
does not necessarily apply in the later stages of oxidation which may only show
small peroxide values.
Determination of Peroxides
A large number of methods is available for the determination of peroxides,
but the results, while reproducible within a given set of experimental conditions,
are difficult to compare from worker to worker as differing experimental
techniques will give discordant values even on the same substrate. The technique
normally involves the liberation of iodine from sodium or potassium iodide in
the presence of peroxide. Lea 63 showed that the system should be acidified
during this release, while Knight TM showed that the presence of other functional
groups did not interfere and Swift 20 showed that one mole of iodine was
liberated by one mole of methyl oleate hydroperoxide. Lea's method 65 of
heating with glacial acetic acid and chloroform in the presence of solid potassium
iodide and a nitrogen atmosphere is claimed to detect as little as 10-6
equivalents of peroxide per g of fat. The reacted mixture when cooled was added
Antioxidants 719
to 5 per cent potassium iodide solution and titrated with 0-002 N sodium
thiosuiphate. A modification which is in common use is due to Wheeler 66 who,
while still using a chloroform–glacial acetic acid solvent (50 ml), used 1 ml of
saturated potassium iodide solution and 3-10 g of the oil under investigation.
Both of these methods, and many other variations, gave conflicting peroxide
values. It has been found that the solvent used and the acid condition will cause
variation, 67 and a modified Wheeler technique using sulphuric acid and identical
sample weights is claimed to have given the most reproducible values. The
importance of constant sample size has been demonstrated by numerous
workers as also has the presence of an inert atmosphere to prevent further
oxidation during the determination 69-71
linoleic acid system have been studied by this method .86 A detailed study of the
accelerated oxidation of hemin in an emulsion at 45°C was made by Berner et
al. 87 The antioxidant was added to the fat prior to emulsification and the oxygen
uptake then measured. The effect of antioxidants was to increase the induction
period. The age and purity of the hemin, pH and peroxide value of the fat and
temperature all affect the induction period. The activity of BHA, propyl gallate,
tert-butyl hydroquinone, tocopherol and synergists (EDTA, ascorbic acid and
citric acid) were studied.
Choice of Antioxidant
The ideal antioxidant should be stable and effective over a wide pH range and be
soluble in its oxidized form, and its reaction compounds should ae colourless and
odourless. Other obvious and essential requirements are that it should be
non-toxic, stable and compatible with the ingredients in the products and their
packages.
The list of effective antioxidants permitted in the USA for use in foodstuffs
includes the materials shown in Table 37.2.
Guaiacum resin
Tocopherols
Lecithin
Propyl gallate
Butylated hydroxyanisole (BHA)
Butylated hydroxytoluene (BHT)
Trihydroxybutyrophenone
Ascorbic acid
Ascorbyl palmitate
Monoisopropyl citrate
Thiodipropionic acid
Dilauryl thiodipropionate
722 Harry's Cosmeticology
Phenolic Antioxidants
Guaiacum Resin
Guaiacum resin is largely phenolic in character, but is a less effective antioxidant
than most of the other phenolics mentioned above. It is more effective in animal
than in vegetable oils and possesses an advantage over some other antioxidants
in that it is equally effective in the presence and absence of water and it is not
seriously affected by heating.
Nordihydroguaiaretic Acid
NDGA shares many of the proprties of guaiacum resin but is more effective
weight for weight. Higgins and Black 12 summarize their studies on the protection
of lard with NDGA as follows: pure lard of low initial peroxide value was
protected against development of oxidative rancidity by 0003 per cent of
NDGA, compared with the requirement of 0 . 006 per cent of propyl gallate. The
stabilizing effect was proportional to the antioxidant concentration over the
range of 0•003-0•03 per cent. A synergistic effect occurred with 0003 per cent
NDGA and 075 per cent citric acid. This, of course, arises from the sequester-
ing effect of citric acid on heavy metals. NDGA is soluble in fats at levels up to
about 005 per cent at 45°C and does not crystallize Out very much on cooling. It
was removed from the US permitted list in 1968.
Tocopherols
These natural materials are not widely used in practice because of their high
price. They have some antioxidant effect with animal fats such as tallow and with
distilled fatty acids, particularly in the presence of a synergist such as citric acid,
lecithin, or phosphoric acid, but are of little value for the preservation of
vegetable oils. Issidorides93 has shown that the action of citric acid with
tocopherols is not only a sequestering effect, but derives also from regeneration
94 gives a method for obtaining
of the tocopherol in the reduced state. Sisley
tocopherols mixed with lecithin as a synergist by extracting wheat germ oil with
dichloroethylene.
Gallates
Gillates constitute one of the most important classes of antioxidants. The propyl
ester is the only one permitted in foodstuffs in most countries but methyl, ethyl,
propyl, octyl and dodecyl gallates are commonly used in cosmetics. Gallic acid
itself is a powerful antioxidant, but tends to turn blue in the presence of traces of
iron.
The valuable antioxidant properties of the esters of gallic acid have been
described by many investigators. Boehm and Williams" found that for practical
and commercial considerations of ease of solubility at low temperatures, acidity
and colour, and from the point of general effectiveness, normal propyl gallate
(that is, normal propyl-3,4,5-trihydroxybenzoate) was the outstanding antiox-
idant amongst the gallic acid esters investigated. These workers found that the
protection afforded by O 1 per cent of the normal propyl gallate in lard was equal
to that obtained by ten times as much Siam benzoin and probably greater than
that of thirty times as much Sumatra benzoin. Peredi showed a tenfold increase
Antioxidants 723
in storage time of lard in the presence of 001 per cent of propyl or ethyl gallate.
However, Tollenaar 97 suggests that gallates are not suitable for the antioxidation
of vegetable oils.
Further recommendation pf the esters of gallic acid has been given by Stirton,
Turer and Riemenschneider. 98 They compared the antioxidant activities of
nordihydroguaiaretic acid (NDGA), propyl gallate, benzyl hydroquinone,
alpha-tocopherol and their synergistic combinations with citric acid, d-iso
ascorbyl palmitate and lecithin in a number of fat substrates: methyl oleate,
methyl linolëate, methyl linolenate and the distilled methyl esters of lard.
Nordihydroguaiaretic acid and propyl gallate surpassed the other substances in
antioxidant activity. Citric acid showed marked synergism with each antioxidant;
the most effective combinations were those of citric acid with nordihydrogua-
iaretic acid and with propyl gallate.
The possible toxicity of ethyl .gallate has been thoroughly investigated, and no
symptoms of toxicity have been observed in mice which received, orally or
subcutaneously, massive doses of ethyl gallate, in concentrations far greater than
could ever be approached by human beings when receiving foods stabilized
against oxidation by the ester. 99
Boehm and Williams95 report that 0•5 g of propyl gallate was administered
orally to one of them on six consecutive days. An examination of the urine
during this period, and for a further six days, showed that no albumen was
present, abnormal sedimental contents were not observed, there being a
complete absence of red blood corpuscles and casts of any kind. They also cite
pharmacological tests received from the Pharmacological Laboratories of the
College of the Pharmaceutical Society (University of London) which concludes
as follows:
1. Normal propyl tnhydroxybenzoate is less toxic than pyrogallol, when ad-
ministered orally to mice (acute tests).
2. There are no acute skin effects observable when a 10 per cent solution of
normal propyl gallate in propylene glycol is left in contact with shaven
guinea-pigs for 48 hours, or the human skin for 24 hours. This contrasts
favourably with the erythematous effect produced under similar conditions by
a 10 per cent solution of pyrogallol in propylene glycol.
Williams'00 gives solubilities of gallates in various oils (Table 37.3).
Non-phenolic Antioxidants
Many non-phenolic antioxidants are chelating agents. Ascorbic acid and ascor-
hyl palmitate appear to act by stopping the free-radical oxidation process.
Anijoxidants 725
Ascorbyl esters are particularly effective in vegetable oils, and make an excellent
synergistic mixture with phospholipids such as lecithin and tocopherol.105
Among the sequestering agents, the thiodipropionatcs are widely used,
usually in conjunction with phenolic antioxidants. The esters of fatty alcohols
have greater solubility in oils.
Monoisopropyl citrate has similar sequestering action to citric acid itself, but a
greater fat solubility.
Lecithin is an effective synergist for many phenolic antioxidants, mainly
because it is an oil-soluble phosphate with excellent sequestering properties.
Members of another class of oil-soluble sequestering agents are. MECSA
(mono-oc.tadecyl ester of carboxymethylmercapto-succinic acid) and METSA
(mono-octadecyl ester of thiodisuccinic acid). Under some conditions these
materials can function as effective antioxidants in concentrations as low as 0.005
per cent—see the review by Evans etal.'°6 They possess the disadvantage that
they decompose on heating, and must therefore be added, like perfume, during
the cooling phase of manufacture..
In general, the effect of any true (that is, chain-stopping) antioxidant can be
enhanced by the proper choice of a suitable sequestering agent to slow down the
initiation of chain reactions at the outset. Citric, phosphoric, tartaric and
ethylenediamine tetra-acetit acids should always be considered as possible
additives to a system that is insufficiently protected against oxidation, before
including more phenolic material. The use of such sequestering agents is
cheaper, and it is less likely to lead to discoloration or odour development, than
the usd of high concentrations of phenols. Of the phenolic materials. BUT is
probably the most universally useful, but each system has its peculiarities which
must he studied at first hand.
Photo-deterioration
Another form of deterioration sometimes encountered is that caused by light in
the visible or the UV spectrum. Such photo-detcrioraion generally manifests
itself as fading of the colour of the product or development of off-colours.
Packaging in opaque containers or wrappers so as to exclude all light is. of
course, an obvious way to avoid this type of deterioration, but this is not always
desirable, nor even necessary. It is frequently possible to wrap or pack in
transparent material suitably coloured or containing a UV absorber to screen
out the offending portions of the spectrum. In some cases where UV energy is
causing the deterioration, the UV absorber can be incorporated in the product.
Mecca was reported 107 as having found that uric acid at 002-05 per cent would
protect solutions coloured with FD&C Blue No.!. D&C yellow No. 10, FD&C
Green No.8 and cochineal, exposed to direct sunlight, while control solutions
without uric acid were completely bleached.
UV-induced deterioration frequently involves the presence of traces of
metals, particularly iron, and when this is the case the screening agent may be
reinforced, or in some cases replaced, by a chelating agent such as ethylene-
diaminetctra-acctic acid (EDTA). The permeability of the cell walls of some
bacteria, notably the Gram-negative Pseudornonas aeruginosa, is altered by
EDTA; Sffith' 011 has found that the addition of concentrations in the region of
726 Harry's Cosmeticology
005 per cent greatly enhances the antibacterial potency of phenolic antiseptics.
Small amounts of this material might thus serve two useful purposes in
protecting systems prone to deterioration by UV and susceptible to the
omnipresent Pseudomonas species.
REFERENCES
36. Myers, J. E., Kass, J. P. and Barr, G. 0., Oil Soap, 1941, 18, 107.
37. Holman, R. T. and Elmer, 0. C., J. Am. Oil Chem. Soc., 1947, 24, 127.
38. Allen, R. R. and Kummerow, F. A., J. Am. Oil Chem. Soc., 1951, 28, 101.
39. Brauer, R. W. and Steadman, L. T., J. Am. chem. Soc., 1944, 66, 563.
40. O'Neill, L. A., Chem. Ind. (London), 1954, 384.
41. Benton, J. L. and Wirth, M. M., Nature, 1953, 171, 269.
42. Wibaut, J. P. and Strang, A., Proc. Koninki. Ned. Akad. Wetenschap., 1952, 55B,
207.
43. Paquot, C. and de Goursac, F., Oleagineux, 1950, 5, 349.
44. Almquist, H. J. and Branch, G. E. K., J. Am. chem. Soc., 1932, 54, 2293.
45. Backström, H. L. 3., Z. physik. Chem., 1934, 25B, 99.
46. Cook, A. H., J. chem. Soc., 1938, 1768.
47. Waters, W. A. and Wickham-Jones, C., I. chem. Soc., 1951. 812.
48. Cooper, H. R. and Melville, H. W., I. chem. Soc., 1951, 1984, 1994. *
49. Sharp, D. B., Whitcomb, S. E., Patton, L. W., and Moorhead, A. D., I. Am.
chern, Soc., 1952, 74, 1802.
50. Bolland, J. L. and ten Have, P., Trans. Faraday Soc., 1947, 43, 201.
51. Davies, D. S., Goldsmith H. L., Gupta, A. K. and Lester, G. R. J. chem. Soc.,
1956, 4926, 4931, 4932.
52. Bolland, J. L. and ten Have, P., Discussions Faraday Soc., 1947, 2, 252.
53. Boozer, C. E., Hammond, G. S., Hamilton, C. E. and Sen, J. N., J. Am. chem.
Soc., 1955, 77, 3233, 3238.
54. Olcott, H. S. and Mattill, H. A., J. Am. chem. Soc., 1936, 58, 2204.
55. Banks, A., J. Soc. chem. Ind., 1944, 63, 8.
56. Lovern, J. A., J. Soc. chem. Ind., 1944, 63, 13.
57. Lea, C. Fl., Research, 1956, 9, 472.
58. Rosenwald, R. H. and Chenicek, J. A., J. Am. Oil Chem. Soc., 1951, 28, 185.
59. .Uri, N., Autoxidaion and Antioxidants, ed. Lundberg, W. 0., New York &
London, Interscicnce, 1961, Vol. 1, p. 133.
60. Fiescr, L. F., J. Am. chem. soc., 1930, 52, 5204.
61. Prosperio, G., Rh'. Ital. Essenze, Profumi, Piante Off., Aromi., Sopeni, Cosmet.,
Aerosol., 1977, 59,424.
62. Lozonczi, B. and Lozonczi, Mrs. B., Olaf. Szappan Kosmit., 1977, 26, 115.
63. Lea, C. H., Food Investigation, Special Report No. 46, Dept. Scientific and
Industrial Research, 1938.
64. Knight, H. B. and Swern, D., J. Am. Oil Chem. Soc., 1949, 26, 366.
65. Lea, C. H. j Proc. R. Soc. (London), 1931, 108B, 175.
66. Wheeler, D. H., Oil Soap, 1932, 9, 89.
67. Nakamura, M., J. Soc. chem. Ind. (Japan), 1937, 40, 206, 209, 210.
68. Stansby, M. E., Ind. Eng. Chem., Analyt. Edn., 1941, 13, 627.
69. Lea, C. H., J. Soc. chem. Ind. (London), 1946, 65, 286.
70. Stuffins, C. B. and Weatherhall, H., Analyst, 1945, 70, 403.
71. Volz, F. E. and Gortner, W. A., J. Am. Oil Chem. Soc., 1947, 24, 417.
72. Jones, J. H., J. Am. Oil Chem. Soc., 1944, 21, 128.
73. Schibsted, H., Ind. Engng. Chem., Analyl. Edn., 1932, 4, 204.
74. Lea, C. H., Ind. Engng. Chem., Analyt. Edn., 1934, 6, 241.
75. Lea, C. H., J. Sci. Fd. Agric., 1952, 3, 286.
76. Wheeler, D. H., Oil Soap, 1933, 10, 89.
77. Becker, E., Gander, K. F. and Herman, W., Fetfe Seifen Anstr..Mittel, 1957, 59,
599.
78. Lea, C. H, J. Soc. chem. Ind. (London), 1936, 55, 293T.
79. Guven, K. C. and Guven, N., Eczacilik But., 1974, 16, 93.
80. Dooms-Goosens, A., J. Pharm. BeIg., 1977, 32, 213.
728 Harry's Cosmeticology
81. Umbreit, W. W., Burns, R. If. and Stauffer,.J. F., Manometric Techniques (Rev.
Edn), Minneapolis, Burgess Publishing Co., 1957.
82. Spetsig, L. 0., Acta Chem. Scand., 1954, 8, 1643.
83. Lew, Y. T. and Tappel, A. L., Fd. Technol., Champaign, 1956, 10, 285.
84. Carless, J. E. and Nixon, J. R., J. Pharm. Pharmacol., 1957, 9, 963 and 1960, 12,
348.
85. Nixon, J. R., Ph.D Thesis, London, 1958.
86. Ozawa, T., Nakamura, Y. and Hiraga, K., Sliokuhin Eiseigaku Zasshi, 1972, 13,
205.
87. Berner, D. L., Conte, J. A. and Jacobson, G. A., I. Am. Oil C/tern. Soc., 1974, 51,
292.
88. Mitchell, J. H., Kraybill, H. R. and Zscheile, F. P., hid. Engng. Chem., Analyt.
Edit., 1943, IS, 1.
8. Chipault, J. R. and Lundberg, W. 0., Hormel Inst. Univ. Minn., Ann. Rpt., 1946,
9.
90. Morris, S. G.. J. Agr. Food Chem., 1954, 2, 126.
91. Williams, H. R. and Moser, H. S., Ana/yt. Chem., 1955, 27, 217.
92. Higgins, J. W. and Black, H. C., Oil Soap, 1944, 19, 277.
93. Issidorides, A., J. Am. chem. Soc., 1951, 73, 5146.
94. Sisley, J. P., Perfurn. Essent. Oil Rev., 1955, 46, 117.
95. Boehm, E. and Williams, R., Q. I. Pharm, Pharmacol. 1943, 16, 232.
96. Peredi, J., Hung. Tech. Abstr., 1955, 7, 24.
97. Tollenaar, F. D. and Vos, H. J., J. Am. Oil Chem. Soc., 1958, 35, 448.
98. Stlrtof), A. J., Turer, J. and Riemenschneider, R. W., Oil Soap, 1945, 22, 81.
99. Hilclitch, T. P. and Lea, C. I-I., Chem. Ind., 1944, 70.
100. Williams, R., Am. Perfunt. Aroniat., 1959, 73(2), 39.
101. Pecreboom, J. W. C., Am. Perfuni. Arotnat., 1959, 73(2), 27.
102. Gearhart, W. M. and Stuckey, B. N., J. Am. Oil Chem. Soc., 1955, 32, 386.
103. Olcott, 11. S. and Kuta, E. J., Nature, 1959, 183, 1812.
104. Knowles. M. E., Bell, A., Tholstrup, C. E. and Pridgcn, H. S., J. Ant. Oil Chew.
Soc., 1955, 32, 158.
105. Lundberg, W. 0., Manuf. Confec., 1953, 33(4), 19, 67.
106. Evans, C. D., Schwab, A. W. and Cooney, R. M., J. Am. Oil Chew. Soc., 1954, 31,
9.
107. Anon, Dru Cos,net. [rid., 1965, 97, 97,
108. Smith, G.. J. med. Lab. Technol., 1970, 27, 203.
Chapter Thirty-eight
Emulsions
Introduction
Every cosmetics laboratory worker knows that emulsions are relatively stable
mixtures of oils, fats and water and are made by mixing oil- luble and
water-soluble substances together in the presence of an emulsifying agent.
Emulsions—creams and lotions—form a very important part of the cosmetics
market; much time is spent in the development of new raw materials by both
suppliers and cosmetics companies. Formulae for good, stable cosmetic emul-
sions are available from books or suppliers' literature and it is not difficult for
even a novice laboratory technician to make a satisfactory emulsion by following
simple written instructions. Clearly, however, no cosmetics chemist can consider
himself competent until he understands how to formulate emulsions of his own
and how to incorporate into them certain desired characteristics. In order to do
this, he must learn something of the fundamentals of emulsion technology. How
far this is pursued will depend upon the inclinations of the individual, but the
aim of this chapter is simply to provide a sufficiently detailed account of these
fundamentals to allow the reader to experiment for himself with understanding
and to form ideas of his own. More detailed information can be obtained from
the references cited at the end of the chapter.
Basic Principles
The starting point in this study is the recognition that certain substances show an
'affinity' for each other and others do not. A simple illustration of this point is
that water and ethanol are completely miscible. Their molecules are quite happy
to exist side-by-side together and show no tendency , to separate into discrete
areas populated largely or exclusively by their own kind. These two materials
show an 'affinity' for each other which is obviously not shared by (say) mineral
oil and water. This idea of 'affinity' plays aji important part in 'emulsion
technology and has to do with the manner in which individual molecules appear
to attract their neighbours in a given environment. Molecules of the same
substance—for example, water—exert an attractive influence on each other and
were it not for the fact that under normal circumstances each molecule is in turn
attracted by many others around it in all directions, any two molecules might be
pulled together. This phenomenon is called 'cohesion' and the force of cohesion
between molecules is attributed to their 'cohesive energy'. The magnitude of
these cohesive forces depends not only upon the size of the molecules taking part
but also Upon their chemical make-up. The basic principle is that 'like attracts
730 Harry's Cosmeticology
like'. Related molecules such as water and ethanol show no tendency to separate
because the cohesive forces between water and ethanol molecules are similar in
magnitude to those between water and water or ethanol and ethanol. When
mineral oil is introduced into water, however, the cohesive forces between water
and mineral oil are negligible compared with those tietween the molecules of the
two materials themselves, and separation rapidly occurs.
'Affinity' manifests itself not only as solubility but also in the concept of the
'phase'. When two or more materials in contact with each other co-exist as
overtly different and separate entities, each is referred to as a 'phase' (Table
38.1). In two-phase systems one phase may be distributed as a large number of
distinct and separate entities in the other. Under these circumstances, the
former is known variously as the 'internal', 'disperse' or 'discontinuous' phase
and the latter as the 'external' or 'continuous' phase. When one material is
dispersed in a finely divided state within another in this way, the area of contact
between the two phases is exceedingly large. It is not surprising, therefore, that
many of the characteristics exhibited by such a system depend primarily on the
chemical or physical natures of the two surfaces and the interaction between
them. This is certainly true of cosmetic emulsions.
Continuous Disperse
phase phase System
Gas Solid Smoke
Gas Liquid Aerosol
Liquid Gas Foam
Liquid Solid Dispersion
Liquid Liquid Emulsion
Solid Gas Foam
Properties of Surfaces
In general, the outermost la yers of materials exhibit very different properties
from the bulk, this being entirely due to the environment in which the surface
molecules find themselves. In Figure 38.1, A represents a molecule some way in
the interior of a liquid. A is surrounded by other molecules which, if close
enough, exert an appreciable attraction on it. These will be contained in a
sphere, Centre A, with a very small but finite radius (represented in the figure by
an exaggerated dotted line around A) .—the 'sphere of molecular activity'. Since
there are as many molecules attracting A in any one direction as attract it in the
opposite direction, there is no resultant cohesive force on A. This is far from
true of a second molecule, B, situated in the liquid surface. Here, forces of
attraction below B are not precisely cancelled out by the attraction of other
molecules above it and a resultant force Fis exerted on B, tending to pull it into
the interior of the liquid. (It should be emphasized that B is very close to the
surface and that Figure 38.1 is greatly exaggerated for clarity.)
Emulsions 731
E
,-
This same consideration applies to all surfaces, whether gas, liquid or solid,
although it is in gases and liquids, in which the molecules have considerable
mobility, that its effect is clearly seen in determining the shape of the material.
Since the inward forces on the molecules in a liquid surface tend to make the
molecules move towards the interior, the surface tends to contract and become
as small as possible, so that the surface area will be the minimum for a given
volume of liquid. Since a sphere is the shape which has the minimum surface
area for a given volume, if no other forces were acting on the liquid, it should be
expected to assume a spherical shape. Relatively large masses of liquid are
subject to proportionately large gravitational -forces; for this reason they will
assume the shape of any retaining vessel (for example, a beaker or measuring
cylinder). For very small masses, however, such as may be found in droplets of
internal phase in cosmetic emulsions, gravitational forces play a relatively minor
role and the droplets may be spherical to a high level of approximation.
Because of the inward resultant force on surface molecules, all surfaces are
said to possess 'surface tension' and the magnitude of the surface tension
depends on F. Of course, if the molecules of material above the surface of the
liquid have little or no measurable attraction for B, then the magnitude of F
depends primarily on the properties of the liquid itself. Should the air, gas or
solid above the liquid surface show some affinity for B, then F and the surface
tension will be lower. Each surface is a boundary or interface whose characteris-
tics, including the surface tension, depend upon the influence of the materials on
both sides.
where T is the interfacial surface tension between the liquid and its surround-
ings. Thus it can he seen that surface tension is no more than the incrasc in
surface energy associated with a unit increase of surface area.
It is a well-known principle in mechanics that an object is in stable equilibrium
when its potential energy is at a minimum. Given the opportunit y , therefore, the
emulsion will lose its considerable energy excess to its surroundings in the form
of heat by coalescence of the droplets of internal phase and phase separation.
(a) By increasing the viscosity of the external phase, he will decrease the
mobility of internal phase droplets making it more difficult for them to
collide with each other.
(b) By ensuring that the internal phase is of the smallest and most uniform drop
size possible, he will decrease the likelihood of adhesion between two drops.
(c) By increasing the mechanical strength of the interface, he will riake this less
susceptible to rupture with the resulting coalescence of adhering drops.
(d) By decreasing the surface tension, he will decrease the thermodynamic
'driving force' for coalescence.
It should be noted that the increase in stability which results from the formation
of internal phase droplets of very small size represents an apparent anomaly. It
has already been shown that decreasing the droplet size causes a rapid increase
in surface area and also that a large surface area can only be achieved, in a given
system, by a larger energy input. Such a system should therefore possess a high
excess energy content—which seems to be in conflict with the rule about high
energy systems being less stable than those of low energy content. Apparently.
therefore, the stabilizing effect of a low probability of adhesion between internal
phase droplets far outweighs the influence of excess free surface energy in
bringing about coalescence.
734 Harry's Cosmeticology
-
S
the resultant inward force, F', is less than before. The surface tension between
the liquid and S is therefore lower than that between the liquid and E. This is
referred to as 'a lowering of surface tension between the liquid and E', although
we can see that this is not strictly accurate. At the same time, the surface tension
between S and E is lower than that between the liquid and E for similar reasons.
Types of Emulsifier
Such is the number and variety of emulsifying agents now commercially
available that their classification must be considered a daunting task—indeed,
there are manuals devoted exclusively to it. Fortunately, the problem of
choosing from among the bewildering variety of products is made easier by the
classification of emulsifiers according to their chemical type and their mode of
action.
Before proceeding to describe the emulsifier types of major commercial
importance, however, a distinction should be made between these (which
function at a molecular level) and certain finely divided solids which have also
been shown to exhibit emulsion-stabilizing properties. Such solids undoubtedly
function by migrating to the emulsion interface, forming a barrier against
coalescence: it follows that the surface of such solids must not be predominantly
water-wettable (hydrophilic) or oil-wettable (lipophilic). Such powders have
little value as cosmetic emulsifiers but since many cosmetic emulsions also
contain suspended powdr'rs (liquid foundations, for example) it is as well to bear
in mind that these could possibly play some part in deciding th stability of the
product.
Figure 38.3 is a diagrammatic representation of a molecule of the more
conventional type of surfactant. The molecule can be considered as being
composed of two parts—a water-loving or hydrophilic group at one end (H) and
an oil-loving or lipophilic group at the other end (L). Since the lipophilic group is
usually a hydrocarbon chain, it is often represented diagrammatically as a 'tail'
as in Figure 38.3. It is easy to see how such a molecule would behave if dispersed
in a single liquid: in Figure 38.4a, the surfactant has been dispersed in oil. Since
the cohesive forces between the hydrophilic portion of the molecule and the oil
molecules is negligible compared to those between hydrophilic ends of the
molecules among themselves, the molecules orientate as indicated in clusters or
'micelles'. The lipophilic parts of the surfactant molecules, experiencing com-
paratively large cohesive forces from the oil molecules, are happy to extend
Outwards into the oily environment.
Figure 38.4b indicates the opposite orientation, encountered when the same
molecules are dispersed in water or hydrophilic media. The same rules apply
here as before except, as would be expected, it is the lipophilic ends of the
I
aT&-
Water
War
Oil
Wa
d e
Figure 3.4 Behaviour of surfactailts in various environments
molecule which cluster closely together in order to escape from the watery
enviroiifllCflt
In Fiiurc 38.4c oil has been added to the water and, as expected, the
surfactant has migrated to the interface. If now an emulsion is formed, each
spherical internal phase droplet will be covered with orientated surfactant
molecules. (Figure 38.4d illustrates a water-in-oil and Figure 38.4c an oil-in-
water emulsion.)
It is eas y to verify . b y experiment, that such interfacial surfactant layers can
stabilize emulsions and this is primarily because the. interface is made less liable
to rupture. To understand this more clearl y , however, it is first necessary to
examine the chemistry of surfaetw in greater detail.
For the hydrophilic end of the molecule the following variations are possible:
1. Introduction of ionizable anionic end groups.
2. Introduction of ionizable cationic end groups.
3. Introduction of amphoteric groups.
4. Introduction of other water-soluble but 'nonionizing' groups such as hydroxyl
or ethoxyl.
All these variations have been used in practice and the classification of surfactant
type depends, customarily, only upon the nature of the hydrophilic end of the
molecule. Emulsifiers may thus be 'anionic', 'cationic', 'amphoteric' or
flO fl i On i C'.
0
/ \ cthylene oxide
Cl 12—C112
Cll(Ct !) ,Cl1 20 it stearyl alcohol
CH1(Cl 1 2 ) 16C1 1 2 (0C}12CH)0l! pol yeth ylene glycol ethers of stear y l alcohol
(i varies between 2 and 30 in most commercial forms). The dotted line indicates
the 'balance point' of the molecule, the groups to the left being oil-soluble and
those to the right being water-soluble. Counting each —CH 2— group or
—CH 3 — group as unity, the lipophilic end of the molecules adds up to 18 and
the hydrophilic end to 3 -i- n (0 = 3; —CH 2CH 20— = 1). If n is less than 15,
therefore, the lipophilic tendencies of the molecule outweigh its hydrophilic
properties, while at values of ii greater than 15 the reverse is true. Moreover, the
greater the difference in numerical value between the two sides, the greater the
imbalance in the relative affinity for the two phases exhibited b y the molecule.
738 Harry's Cosmeticology
If it = 20
total molecular weight = 253 + 897 = 1150
The 1-ILB value can only be determined in this simple way for nonionic
emulsifiers of known and definite composition. The concept is also applicable to
anionic or cationic emulsifiers, although it is possible to exceed the theoretical
upper limit of 20 with these materials. This does not detract from the practical
merits of the l-ILB system, but it does mean that alternative methods of
determining HLB values have to be used. This will be described later.
Figure 38.5 summarizes the applications of surfactants of various HLI3
values.
0 5 10 15 20
HLB value
Figure 38.5 Practical applications of surfactants of various ULB values
A Detergents
B Solubilizers
C Surface wetters
D Water-in-oil emulsions
E Oil-in-water emulsions
than 10), then T 0 is larger than T. Remembering that the net inward force, F.
on molecules in a surface is directly proportional to the interfacial surface
tension, the surfactant surface has it propensity to curve towards the side having
the. greater surface tension—in this case, towards the oil phase 2 - With
mechanical agitation, therefore, such a system would produce overwhelmingly
an internal phase of oil dispersed in water. For surfactants having it low l-1L13
value, exactly the reverse is true and a water-in-oil emulsion is the most probable
end result.
Thus it can be seen that the relative lowering of the surface tension on each
side of the surfactant interface helps to determine the nature of the emulsion and
the ease of emulsification. It does not, however, determine the stability of the
emulsion—this is a point which deserves great emphasis. The essential factors
governing the integrity of the interfacial film and its resistance to rupture are its
extension, its compactness and its electrical charge .4 A towering of interfacial
tension is not vital for the stability of an emulsion.
Factors Contributing to the Strength of the Interfacial Surfactant Fil,n
Perhaps the most obvious requirement is that there must be sufficient emulsifier
present to form at least a monolayer over the surface of the internal phase
droplets—and this, in turn, will depend upon the droplet size. 56 It is easy to
verify in the laboratory that even the most stable cosmetic emulsion can be
rendered unstable by progressive reduction of the quantity of emulsifier used in
its production. Indeed, stability can usually he improved by it certain excess of
emulsifier over this essential minimum. Evidence exists' that when there is a
danger of a break in the interfacial film, the interfacial tension at the threatened
point rises and signals to the reserve emulsifier molecules in the continuous
740 Harry's Cosmeticology
phase that there is a fault. These spare emulsifier molecules close in and repair
the damage. It appears that mixed interfacial films (films formed from more than
one emulsifying agent), are able to resist breakage in this manner even more
easily.
Until now it has been assumed that, given sufficient emulsifier, there is no
impcdimeiit to the formation of a compact, continuous film of emulsifier
molecules in the interface. This is far from true, however. Not onl y do these
identical O1OkCUICS occupy a monolayer, but they are orientated so that
neighhouriii molecules have identical parts of their structure in close proximity.
As a consequence of the balance of forces between like molecules at ver y close
proximity. such a tightly packed monolayer is not thermodynamically stable, and
the emulsifier molecules are constrained to remain separated from each other, so
weakening the strength of the interfacial film. Where the hydrophilic ends of the
molecules are ionized (anionic or cationic emulsifiers) this separation and
weakening is further exaggerated by the repulsion of juxtapositioned like
electrical charges.
The second effect tending to interrupt the contrflUity of the interfacial
monolayer is simple steric hindrance. Consider, for exampiv., the problem
experienced by neighbouring molecules such as the polyoxyethylenc sorhitan
unsaturated esters indicated in Figure 38.6. Obviously it would not be easy for
such molecules to pack closely together in an orientated monolayer. The
importance of these steric effects call be judged by the fact that whereas
soaps of monovalent metal ions tend to form oil-in-water emulsions, the similar
soaps with pol y valent metal ions produce largely water-in-oil emulsions. In this
case, steric hindrance actually dictates the direction of curvature of the interface.
One way of overcoming these problems is to incorporate one or more
additional species of emulsifier molecule into the interfacial film—in other
words, to use a mixed emulsifier system. It is relatively simple to show,
experimentally, that mixed emulsifiers produce more stable emulsions from a
given oil-water mixture than single emulsifiers — provided only that the emul-
Unsaturated
\\ hydrocarbon
Polyoxyethylene
sorbitan
siflrs chosen are chemically and ph y sicall y compatible with each other. The
choice of a combination of anionic and ciii I1)liic molecules. for example, would
I unwise because of chemical conibi nation and elect rochemical neutralization
of charges. This does not mean. Iio e Cr. that the y have to be of similar HLB
satucs In fact, better results are obien achiesed h a combination of surfactant
molecuics having widely differing EILB values combined in such quantities as
to produce it rcsu It ant Ii LB close to the opt iiii urn for the s St C fl to be e mu isi lied
this latter point will be discussed not e 111hy later) . I lie reason for this
phenomenon is clearly that the IWO (liSSitllil,it illi dccular t\ pCS can. b\ alternat-
ing in the interface, form a much more closets packed. condensed interfacial
film. In Figure 38.7a the iniernal phase has been surrounded b y an interface of
of steric hindrance
surfactant molecules of low 111,13 value which, because
and the mutual repulsion of che mica Iv identical entities. Ii as fo ruied only a
discontinuous film. Figure 38.7b shows this surtactimnt partially replaced by two
others with widely different HLB values and chemical composition allowing it
close-packed, continuous film to be produced which acts as a much better
mechanical barrier to coalescence. Sometimes the second emulsifier can have
s
ver y little emulsifying potential at all--that is, it can have an Ill. B value clo e to
zero—just so long as it can migrate to the surface and 'insulate' insulate' the other
molecules from each other. Cetyl and oleyl alcohols will function in this manner
and so will glyce r y l monostearate
Water m
oil______
•0
a b
Figure 38.7 Effect of mixed emulsifiers on crnulsion stability
a Poor stabilit y b Good stability
I Lovk 11L.13—unsaturated
• Medium FILB—unsaturated
0 high
742 Harry's Cosmeticology
More surprisingly, it has been shown that charges at the interface also OCCU
W hen a totally nonionic emulsification system is used. This has been accounted
for b y frictional forces originating in the movement of oil droplets in the
cOntinuous phase. It has been shown that electrical charge can result when two
liquids with different dielectric constants are mixed and that the one with the
higher dielectric constant it-.alwavs positively charged while the one with the
lower dielectric constant is alwa y s negatively charged. (Thus in oil-in-water
floflionic emulsions, the oil droplets are always negatively charged.)
Although no method exists for measuring surface potentials in an emulsion
directl y , the zcta potential' can readil y he evaluated by measuring the velocity
of charged droplets in an applied DC field. The measured values of emulsion
zeta potentials for nonionic emulsifier combinations have proved to he sur-
prisinclv high (about 40 mV) perhaps even more interesting is the fact that,
when zeta potential is plotted graphicall y against the HLB of the emulsifier
conihination, the maximum zeta potential is always found to coincide with the
optiiiiuiii 11 1.13 value for the particular s ystem studied.
If mobile ions are present in the external phase of an emulsion, they are
attracted by the charged droplets of the internal phase (if these have an opposite
charge) giving rise to the formation of an electrical double layer. The nature and
effect of this double layer arc markedl y different in oil-in-water emulsions and
water-in-oil emulsions (Fi g ure 3i). The thickness of the double la yer around
Oil droplets In oil-in-water emulsions aFiloulits to only 10 to 10 m.
Electrical repulsion therefore occurs at very short inter-globular distances and
this results in a very consicrahle electrical harrier which must be overcome
before two droplets can coalesce. Oil 01 her hand, the electrical double layers
around water droplets . in oil are very diffuse (several m in size) and the
electrical potentials of adjacent droplets overlap, lowering the potential harrier.
The stability of water-in-oil emulsions cannot, therefore, be attributed to
electrical repulsion of char ged droplets.
01W
d -
Fiztirc 35.8 Potential energy (f) of repulsion due to electrical double layers plotted
agiwist distance (d) between internal pliae droplets
Emulsions 743
hindrance, prevent droplets coming together. Both these phenomena have a
stabilizing influence on the emulsion but there are other factors which may also
affect the stabilit y for better or worse.
The simple picture given by Stokes's Law shows that mobility of internal phase
droplets is affected by the viscosity of the continuous phase, the difference in
density between the oil and water phases and the size of the disperse phase
droplets. Of these, the differential density between the two phases is least
amenable to experimental control, although it is obvious that the smaller this
difference is, the less is the likelihood that the internal phase will float to the
surface or sink to the bottom of the emulsion.
Temperature
In has already been stressed that optimum stabilit y is achieved by the correct
choice of emulsifier combination. The chosen emulsifiers must he compatible, of
correct HLB value and of correct chemical type. These last two characteristics
are heavily dependent upon the relative solubility of the hydrophilic and
lipophilic ends of the surfactant in the water and oil phases respectively.
Solubility, however, is very temperature-dependent. It is unlikely that as the
temperature of an emulsion changes, the relative solubilities of both ends of all
its emulsifier system will change in strict proportion. In other words. 1-11-I-3 is to
some extent if property itself. Variation of temperature
can therefore decrease the stability of all This is obviously something
which must be borne in mind when formulating products for differing climates.
ilowever, the stability of emulsions is often tested in the laboratory by storage at
elevated and refrigerator temperatures (and sometimes by cycling between these
two extremes) fit hope that such methods will give a rapid indication of the
stabilit y oil storage at average temperature. Such a practice is highly
questionable. The stabilit y testing of emulsions is discussed more fully later in
the chapter.
Concentration Of ions in the Water Phase
The dielectric constant of the oil phase of an emulsion is not great enough to
allow an y ionizable chemical species to dissociate to any great extent and it has
alread y been seen that this reduces the stabilizing influence of an\ electrical
double la y er when oil is the continuous phase. Dissociation is an important
factor in oil-ill-waterr emulsions, however, not onl y becaLisc of the influence on
ionic emulsifiers but also because of the effects of other soluble ionizable species
in solution including hydrogen and hydroxide ions.
The pH value of emulsions is if which is often discussed for a variety
Of reasons, not least because of its effect on stability. It is necessary to remind
ourselves, however, that p!-1 is if of hydrogen ion activity in an aqueous
environment. Whether the term should strictl y be applied to an oil-in-water
emulsion is a moot point since the influence of the oil phase and emulsifiers on
h y drogen ion activity is probabl y unknown. However, common usage dictates
that the measurement of the pH of emulsions, when the continuous phase is
aqueous, will continue. Under no circumstances call concept of pH tic
applied when oil is the continuous phase.
Nevertheless, the influence of h y drogen ion concentration in oil-in-water
emulsions is dramatic 's hcnever an ionizable emulsifier system is used because of
the change at species which can he brought about. Anionic emulsifiers are
converted to non-ionizable salts in acidic media and the reverse is [rue of
cationic emulsiliers. In both instances, water solubilit y and therefore all emul-
sifier activity can he lost. The effect of p!-I oil emulsifiers is less
dramatic but obviously dictates whether the anionic or cationic form predomi-
nates.
The presence of other mobile ions in solution in the external phase also has an
important influence. Thev are attracted by the charged droplets of the disperse
phase. as we have seen. The zeta potential falls as more electrolyte is added and
this can decrease emulsion stability. Even at relatively high electrolyte concert-
Emulsions 745
trations and zeta potentials close to zero, many emulsions remain stable. This
must be due to steric stabilization and the mechanical barrier afforded by a good
interfacial film.
are made in precisely the same way. Usually, one or more of the emulsions will
give better stability than the others. Should they appear to he uniformly good,
however, the experiments must be repeated using less emulsifier; more emul-
sifier should be used if the set is uniformly bad. Occasionally, two combinations
having widely different HLB values show outstanding ability. In this case the low
value probably relates to a water-in-oil and the high value to an oil-in-water
emulsion. This trial-and-error process has now enabled the experimenter to
arrive at an idea of the optimum HLB value for his system. At this stage, a more
accurate determination can be achieved by producing a second set of emulsions
using the same emulsifier pair but combined to give HLB values in smaller
increments close to the value obtained in the first series of experiments. For
example, if the initial value was found to be 8, then a series 74, 76, 78, 80, 82,
84 might be used.
Orientation of Phases
Three factors combine to determine which phase will be continuous and which
phase will-be disperse in a cosmetic emulsion: the type of emulsifier system used,
the volume ratio of light to heavy phase and the method of manufacture. All
three factors are interrelated, but any one of them can exert a controlling
influence—at least, during the initial formation of the emulsion. It is well
known, however, that spontaneous changes in the orientation of phases can
occur—a phenomenon known as 'phase inversion'.
The effect of the HLB value of a chosen emulsifier system has already been
discussed. Clearly, the phase having the greatest interfacial surface tension tends
to produce a concave surface so that, other factors allowing, it becomes the
internal phase of the emulsion. If the surface tensions on both sides of the
interface are equal—or nearly so—then inversion might be expected to take
place readily. It has been shown that there is an HLB value at which inversion
takes place most easily. 16
Emulsions 747
Volume Ratio
Theoretical calculation shows that the maximum volume which can be occupied
by uniform spherical particles is 74 per cent of the total liquid volume.
Emulsions may, however, be prepared with internal phases amounting to 99 per
cent of the total liquid volume. 17 This is possible because the spherical droplets
can become distorted in shape. In such emulsions, the internal phase particles
become increasingly angular as they are crowded together. Not surprisingly, the
effect of high internal phase concentrations is to produce emulsions of greatly
increased viscosity and their ability to remain stable depends primarily on the
mechanical strength afforded by a highly condensed and structured interfacial
surfactant film.
Method of Manufacture
Although there are exceptional cases, it is generally difficult for a phase to be
dispersed in a stirred tank if it occupies more than 74 per cent of the total liquid
volume. Either liquid may, however, be dispersed over a wide range of relative
volumes (the ambivalent region) and for systems containing no emulsifier the
choice of dispersed phase is often dependent on the manner in which dispersion
is initiated. If a simple, two-layer mixture of water and oil phases is agitated it
will tend to form an oil-in-water system if the agitator is sited in the water phase
or a water-in-oil system if the agitator is immersed entirely in the oil phase,
because the disperse phase is most likely to be the one that is drawn into the
other. For the same rcson, if the vessel is initially filled with one phase (prior to
the addition of the second phase), this initial phase will be the continuous one.
It appears that the orientation of the emulsion is also affected by the type of
agitator used and its speed. For a given stirrer speed, there exists a volume ratio
(of light to heavy phase) above which the heavy phase is dispersed, and a region
at lower volume ratio below which the lighter phase is dispersed. Between these
limits lies the ambivalent or metastable region where either phase may be
dispersed—but this shows a strong hysteresis effect. Thus if, at a constant stirrer
speed, water is added to a stable water-in-oil emulsion the system will eventually
invert at the lower volume ratio limit. If oil is now added to this emulsion,
reinversion will not occur until the ambivalent region has been traversed and the
upper limit reached. This ambivalent region can be increased by adding solutes
which are partially soluble in both phases, such as surfactants.
As the stirrer speed is increased, the inversion points of all volume ratios tend
to increase asymptotically to a constant value which is dependent upon the
stirrer design. It has also been noted that for equal phase volumes at high stirrer
speeds, the heavy phase tends to be the continuous one.
Another interesting point is that droplet size of dispersed phase varies with
volume ratio and is greater when the lighter phase is dispersed. Thus, on
inversion, a step change in droplet size occurs.'8
These observations provide the background to various methods of practical
emulsion manufacture.
I Charged liquid
adensed
::t
¶!
Oil Oil
b Multilayer association
- Water torces.Water squeezed
out on shaking
Oil - Oil
Oil
C WaterRigid neutral
Oil cannot / film
Obe squeezed out
Figure 38.9 Schematic rcplcscntatii)fl of phase inversion
o Na eels'! sulphate
• Cholesterol
Emulsions 749
the energy change which brings inversion about has to do with flow and
turbulent behaviour.
Except when used purposely, phase inversion is not often encountered in
cosmetic production provided that the production chemist is aware of the
conditions under which it can occur. 16.19 It is probably true to say that at the
present time inversion cannot always be predicted but that the conditions which
increase the risk can be recognized and it can be explained when it does happen.
Characteristics of Emulsions
Having discussed the factors affecting the stability of cosmetic emulsions in some
detail, it is appropriate to turn to the other characteristics by which they are
judged by the user and the means byw hich these can be controlled.
Of prime importance when considering cosmetic emulsions is their appear-
ance, since this can help to determine their customer appeal. Emulsions may
var y tremendously in appearance from glossy opaque whiteness through a grey
ti :!uccile sparkling clarit y . Opacity is due to two interrelated factors: the
Emulsions 751
size of the internal phase droplets and the difference between the refractive
indices of the internal and external phases. Light is reflected and refracted at
each interface between droplet and continuous phase. Such changes in direction
are so numerous (because of the large nuniber of droplets) that much of the light
escapes from the emdlsion surface in the same direction that it entered— that is,
back towards the viewer. If, however, the refractive indices of both phases are
identical, or nearly so, no such reflections and refractions take place; light
travels unhindered through the emulsion which has a sparkling clear appear-
ance. This applies no matter what the size of the internal phase droplets. If the
droplets are large, however, each ray of light encounters only a small number of
interfaces during its passage through the emulsion. Sufficient light is reflected
back towards the viewer to make the presence of the droplets obvious, but the
bulk of the light, which is refracted, can find its way through. This accounts for
the globular appearance of emulsions in and advanced stage of aggregation and
separation. As the particle size of the internal phase diminishes, the familiar
milky-whiteness appears: as the size reduction continues, the colour takes on a
blueish hue, becoming grey, semitransparent and finally transparent. These
changes in appearance occur as the particle size of the droplets approaches that
of the wavelength of light itself. The probability that a light ray will collide with
(and he reflected by) a tiny particle is enormously reduced once the particles
become so small that they are comparable in size to the wavelength of light.
Under these circumstances the majority of rays pass through the emulsion
without being reflected or refracted and the emulsion appears to be transparent.
As the droplet size approaches that of the wavelengths at the red end of the
spectrum, the reflected or refracted light is made up increasingly of smaller
wavelengths at the blue end of the spectrum until, eventually, the droplets
become too ,mall for interaction at all;
In practice, it is difficult to formulate emulsions in which both phases have
similar refractive indices—micro-emulsions arc far more frequentl y found,
although even these are not common (Table 38.2).
The gloss of the emulsion is a function of the microscopic smoothness of its
surface. For ultimate smoothness and gloss, the internal phase particles must be
relatively small and even in distribution and there must he no inclusions in the
external phase such as large crystallites of stearic acid or inorganic matter of
large particle size.
Internal phase
droplet size Emulsion appearance
0•5 mm Globules clearly visible
05 mm to 1 j.tm Milky-white
1 jm to 0 . 1 14M Blue-white
01 Am to 005 Am Grey, semi-transparent
<005 Am Translucent or transparent
—49
752 Harry's Cosmeticology
R/ieological Properties
The rhcological behaviour of emulsions is an important subject, not only
because of its influence on the 'feel' and acceptability to the consumer but also
because of its impact on the manufacturing process. The science of rheology
relates to matter which is being deformed or made to flow by applied forces.
Figure 38.1() represents an emulsion flowing steadily under a constant force
through a pipe. The layer A in contact with the pipe is practically stationary but
the central part C of the emulsion is moving relatively fast; in this sense, the
emulsion is not only flowing, but is being deformed. At other lyers between A
and C (such as B) the emulsion has a velocity less than at C, the magnitude of the
velocities being represented by the length of the arrowed lines in the figure.
Since the velocities of neighbouring layers are different, a frictional force is
generated between them just as in the case of two solid surfaces moving over,
each other.
It was Newton who first suggested that this frictional force, F, was propor-
tional to the area of surface considered and the velocity gradient in the part of
the liquid at the point of interest. Thus:
F = iA X velocity gradient
where .4 is now an area and the proportionality constant, i, is known as the
coefficient of viscosity.33
When materials are subjected to deformation of the type illustrated in
Figure 38. 10, the y are often said to he under the influence of 'shearing' forces.
Thus the quantity F/A is 'iiown as the 'shear stress' per unit area. To be
consistent, the velocity gradient is similarly referred to as the 'rate of shear'. In
this way, the viscosity of emulsions and other liquids can be defined as the shear
stress divided by the rate of shear:
F'
7=.
A
Figure 38.10 Different flow velocities of layers of emulsion in a pipe
Emulsions 753
'Newtonian' behaviour and include water, hydrocarbon oils and certain other
liquids such as low-viscosity silicone oils. Many liquids, however, including the
vast majority of emulsions, show deviations from this simple model and are
therefore to be regarded as 'non-Newtonian'.
In Figure 38.11 curve A illustrates the relationship between viscosity and rate
of shear for a Newtonian fluid; curve B illustrates the case where the viscosity
apparently falls as the rate of shear increases. In some instances, a certain force
must be applied before any shear (or flow) takes place at all—such materials,.are
said to exhibit 'plastic' behaviour. Many fluids, on the other hand, merely
exhibit a decrease of viscosity as rate of shear increases from zero, and these are
referred to as 'pseudoplastic'. Dilatant materials 'firm up' as rate of shear
increases; emulsions of this kind are infrequently encountered. Much more
common are emulsions which show a degree of 'thixotropic' behaviour. Thixo-
tropic materials exhibit reversible behaviour—in other wordr, after a lowered
viscosity caused by increased rate of shear, a subsequent eduction of shear
results in a corresponding increase in apparent viscosity. TFis increase may not
be immediate, and recovery can be slow. The majority of emulsions are
non-Newtonian and show some degree of thixotropic behaviour, although the
complete recovery of the initial viscosity is not always achieved.34
Two factors contribute to the viscosity of emulsions: the viscosity of the
external phase—which has been dealt with already—and the ratio of internal to
external phase. Apparent viscosity increases with the proportion of internal
phase. In extreme cases, where this exceeds 74 pc cent of the total volume, the
emulsion can be transformed so thth it has a paste-like consistency and
examination under the microscope shows that The usual spherical shape of the
internal phase droplets has become angular and distorted. 17 Such viscosity is
referred to as 'structural' viscosity and may he achieved with emulsions of either
type.34
Finally, it should be noted that air trapped in the emulsion can cause a
considerable increase in apparent viscosity, particularly if it is very finely
divided.
S
Figure 38.11 Viscosit y (j) versus rate of shear (S) for a Newtonian fluid (A) and for a
pscudoplastic material (B)
754 Harry's Cosmeticology
Application Properties
The in-use properties of Cosmetic emulsions can be thought of as those which are
apparent during its application to the skin or hair (the 'feel') and the after-effects
once the product has been applied. Both types of property are important since
even the most effective products will not appeal to the consumer if the 'feel' on
initial application is unpleasant.
The initial feel of an emulsion is largely dependent on that of the external
phase; thus an oil-in-water emulsion will feel like water, whatever is dispersed in
the aqueous phase. Water-dispersible thickeners and additives such as glycerin,
sorbitol and glycols will all exert some effect. Water-in-oil emulsions will feel
oily—but whether or not they are sticky, for example, depends upon the choice
of oil--phase ingredients. Viscosity also plays an important part in the initial
impact of an emulsion: high viscosities tend to give a cream 'richness'.
During the application, some emulsifiers tend to promote the appearance of a
foam-like whitening, often referred to as 'soaping'. Anionic emulsifiers are
particularly prone to this and the effect is not always considered advantageous
since it prolongs the application time.
As the water and other volatile ingredients evaporate, the 'feel' changes.
Oil-in-water emulsions invert; this may happen abruptly or gradually but the
difference in 'feel' as this happens, while being easy to detect, is somewhat
difficult to describe in words.
Finally, the after-effects are determined by the choice of oil-phase ingredients
(which may be greasy or non-greasy) and any non-volatile water-phase ingre-
dients.
Apparent Viscosity. Several methods of determining this property are used in the
cosmetics industry. Better still is the measurement of apparent viscosity at two
rates of shear (preferably with a ten-fold different in shear rate) since the ratio of
the two readings gives an indication of the degree of non-Newtonian behaviour
(Figure 38.11).
Chemical Identify of Separated Phases. This may be required for the purposes of
evaluating emulsions of unknown composition or to check that certain key
ingredients (for example, preservatives) have been added to the emulsion under
examination. This is a rather specialized topic and is normally reserved for the
expert analytical chemist, although analysis of this type is occasionally reported
in the general literature.
REFERENCES
Introduction
It is probably true to say that processing in the cosmetics industry has evolved
largely through practical experience and principles gleaned b y analogy with
other industries rather than by many fundamental studies. While the literature
abounds with reports from workers in the areas of product development and
efficacy, little seems to have been reported on new production technology over
the last few decades. This does not necessarily indicate that cosmetics manufac-
turers see no need for improving their production facilities; on the contrary.
especially those associated with relatively large-scale
production problems —
processes _continue to cause difficulties and delays even in the largest and best
equipped factories.
The greatest single obstacle to process improvement in most plants is the
enormous variety of product types, each with its own set of physical and
chemical characteristics, which must be dealt with during the course of .a veal.
The need for flexibility is usually of great importance and this leads towards
compromise and except in the largest manufacturing units, away from equip-
mnt especially designed to perform specific tasks.
It is particularly important. therefore, that chemical engineers and production
chemists in the cosmetics industry should understand the basic principles and
characteristics of the plant at their disposal and that they should he sigilant in the
search for new equipment which will perform with even greater efficiency the
tasks for which they arc responsible. broad range of
While cosmetics manufacture is concerned with a ver y
processes, there are enough common elements to allow ;I simple
overall view of the subject; this helps considerably in a study of the basic
principles of cosmetic production technology. -
The first step in this simplification procedure is the division of the subject into
two parts: bulk manufacture and unit manufacture.
Not only are all these operations different from each other, but at each stage
the characteristics of the bulk are quite different and require a different set of
processing characteristics to achieve the optimum economic process. Not sur-
prisingly, therefore, the optimum is rarely achieved.
FIlL subject of pumping is not clearly separated from that of mixing since
pumping implies the forced flaw of product. Any flow will naturally introduce an
element of mixing if the product is not already homogeneous. Further, since flow
The Manufacture of Cosmetics
759
is a common element of both processes, the same product characteristics (for
example, theological behaviour) must be taken into account.
Filtering is not usually a Unit operation of major importance in cosmetics
manufacture except in the production of spirituous preparations (colognes,
aftershave -ind perfumes). It is possible to regard filtering as un-mixing and
certainly the flow characteristics of the filtered product are again of prime
importance. The use of sub-micrometre filters for the sterilization of water is
discussed elewhere in the book.
Unit Manufacture
Most cosmetic products are filled from bulk in machines specifically designed to
handle the units of a particular product type. While it is true that great care must
be taken in the choice and the setting up of such machines, the main problems
encountered are often concerned with the characteristics of the machines
themselves rather than with the manufacture or processing of the product. There
are at least two areas, however, where special understanding of the product units
and their characteristics are essential for the achievement of efficient production:
these are the moulding processes (lipsticks, wax-based sticks, alcohol–stearate
gels) and compression processes (compressed cyeshadow, blushers and face
powders).
A description of unit manufacture could include all filling and packaging
operations; for the purposes of this chapter, however, discussion will be confined
to bulk manufacture.
Definition of Terms
The object of a mixing operation is to reduce the inhomogeneities in the material
being mixed. As Table 39.1 shows, inhomogencity may be of physical or
chemical identity or of heat. Further, in the processes demanded by cosmetic
manufacture, the mixing is designed to be permanent—or as permanent as it is
possible, to make it—as distinct from those operations (such as extraction and
Ipping) which rd' on eventual un-mixing in order to achieve the desired
objective.
Clearly, the degree to which in homogeneity can be reduced depends on the
efficiency of the mixing apparatus used and also on the physical characteristics of
the materials Constituting the mixture. For miscible liquids, homogeneity can be
produced at a molecular level whereas for mixture of powders homogeneity is
limited to the sizes of the powder particles themselves. When examing a mixture
for quality, therefore, the scale of scrutiny—the magnification at which the
mixture is exa mined—must vary from product to product. At an acceptable
scale of scrutiny: perfect mixing implies that all samples removed from the
mixture will have exactly the same composition. This is rarely achievable.
Random mixing
is achieved if the probability of finding a particle of a given
component in a sample is the same as the proportion of that component in the
whole mixture. Random mixing is the aim of all industrial mixing operations and
Harry's CosmeticologY
760
whereas samples removed from such a mixture will not be identical, the
vat iations should be very small. If the scale of scrutiny is reduced sufficiently,
however, this may no longer be true.
Mixing can only occur by relative movement between the particles of the
constituent components of the mixture. Three basic mechanisms for achieving
this relative movement have been identified: bulk flow, convective iixing and
(which includes shear mixing, cutting, folding and
diffusive mixing. Bulk flow
tumbling) occurs in pastes and solids, when relatively large volumes of mixture
are first separated and then redistributed to another part of the mixing vessel.
Convective mixing
involves the establishment of circulation patterns within the
mixture. Finally, diffusive mixing occurs by particle collisions and deviation from
a straight line. In miscible liquids of sufficiently low viscosity, the thermal energy
which is possessed by the constituent molecules may be enough to achieve a
good mixture quality by thermal diffusion without additional energy .being
applied, although this process is usually too slow for industrial purposes.
It is incorrect to assume, however, that the relative movement between
mixture particles brought about by these mechanisms always results in an
improved mixture quality; on the contrary, many mixing problems arise from
the tendency of mixture particles to segregate during attempts to mix them.
Segregation is defined as the preference of the particles of one component to be
in one or more places in a mixer rather than in other places. The size of the
non . uniformitics in an imperfect inmixture is sometimes referred to as the'scale of
composition between neighbouring lumps or
segregation' and the difference :
volumes is the 'intensity of segregation'. Segregation is not, fortunate[), n.jor
problem in cosmetics manufacture although it does manifest itself occasionally
(as, for example, in the flotation of pigments during lipstick processing).
SOLID—SOLID MIXING
Table 39.1 distinguishes between two types of solid—solid mixing operation:
those concerned with segregating powders and those with non-segregating or
cohesive powders. The essential difference between these two categories relates
to the properties of the powders themselves and, in particular, to the freedom
which individual particles have to move independently of their neighbours.
Free-flowing powders exhibit many process advantages (such as easy storage,
easy flow from hoppers, smooth flow of product), but have the disadvantage that
they tend to segregate unless all the constituent particles are of very similar
shape and size. Cohesive powder, on the other hand, lacks mobility, and
individual particles are bonded together and move as clumps or aggregates.
Although segregation does not appear to be a problem (except, as will be seen,
and do
at ver y small scales of scrutiny), cohesive powders are difficult to store .
not easily flow from hoppers.
In a powder mass, there are forces at work which tend to make the particles
bond to each other and these are balanced by the gravitational masses of the
particles which cause them to fall apart again. Although the bonding forces, for a
given powder. are largely independent of particle size, their gravitational mass is
obviously not. For this reason, particles will stick togcher only when they are
The Manufacture of Cosmetics 761
small enough for the gravitational forces acting on them to be much smaller than
the bonding forces. Powders composed primarily of such particles exhibit
cohesive characteristics and those consisting of larger particles tend to be
free-flowing. To a first approximation, therefore, the division between the two
types of powder is one of size and the critical size is approximately 50 m: below
this particle size, powders are cohesive.
Figure 39.1 shows the particle size range of commercial grades of some
powders commonly used in cosmetics production; by inference it will be noted
that they are all predominantly cohesive in nature.
Figure 39.1 Range of particle size of some powders commonly used in cosmetics
A Titanium dioxide
B,C Magnesium carbonates
D Mica
E Zinc stearate
F,G,i-1 Micas coated with titanium oxide
I Talc
J Organic pigments
K Inorganic pigments
762 Harry's Cosmet:coiogy
3. The forces are very much weaker than those holding the particles themselves
together; that is, it is much easier to break up agglomeraxcs than it is to break
up a primary particle.
4. The probability of a small particle bonding to a larger one is much greater
than that of a particle bonding to another of the same size.
pd hopper
T ;crew feed
outset
Figure 39.2 Hammer mill
The Manufacture of Cosmetics 763
which the hammers move (60-100 in increases the chance of a hammer
making contact with each particle and the dwell-time of particles within the
chamber is increased by the placement of a variable size screen over the exit.
Hammer mills are very efficient in the comminution of brittle particles in the
range of 1500-50 p.m but below this size their efficiency (the probability of direct
impact) falls off rapidly. This is fortunate, since it means that cosmetic talcs and
micas can be passed through without being substantially altered. At the same
time, however, the very high rotational speed of the hammers and the air flow
within the chamber ensure that there are enough weak secondary impacts
(particle–wall and particle–particle) to break the much weaker pigment
agglomerates—which may be up to 50 p.m in diameter. The disintegrated
agglomerate fractions then stabiliie by becoming coated on to larger talc
particles and should not be further changed by subsequent passes through the
mill.
Nevertheless the hammer mill, in its role as an extender of pigments Onto talc,
has certain disadvantages. For example, most of the extremely high energy
which it makes available is wasted and is largely dissipated in heating the
powder. From the viewpoint of energy consumption, therefore, a hammer mill
used in this way is very inefficient. The feed-rate and therefore the processing
time for all but the smallest batch sizes of powder is very slow. Attempts to
speed up the process by the substitution of exit screens of large diameter often
result in inadequate extension, necessitating a second or third pass through the
mill. On the other hand, increasing the residence time of powder within the
grinding chamber by decreasing this mesh size can cause the screen to become
blocked with compacted powder, resulting in overheating and damage to
machine and product. Perhaps the biggest disadvantage of all, however, is that
the hammer mill is a continuous processing device being used for batch
processing. For this reason, it must be fed with a powder mixture which has
already been effectively mixed, otherwise the colour of the milled product
changes as each Section of unmixed bulk passes through. This preliminary
mixing must be efficient although it is not necessary for any extension to be
achieved at this stage.
Since the pre-mix is an additional operation and adds to the processing time,
the mixer which is selected must he as efficient as possible. Probably the most
widely used is the ribbon blender' which comprises a horizontal drum contain-
ing a rotating axial shaft which carries ribbon-like paddles. In such a device, the
pre-mix can take anything between 20 and 60 minutes. Other mixers are now
available which utilize higher energy input but are quicker. Table 39.2 summa-
rizes the properties of some of the more conventional powder mixers. Since it is
relatively easy to achieve good mixture quality (at a large scale of scrutiñy)in
cohesive powders, any mixing device will eventually producéa satisfactory even
distribution of components provided that it contains no dead spots where mixing
does not take place.
It is usual to add the liquid binder during this preliminary mixing stage. The
binder may be poured into a suitable orifice in the mixer although many
production chemists prefer to spray it into the mixer cavity as an aerosol through
a venturi or simihir device. This procedure helps to distribute the liquid more
even! and avoids the formation of wet, lumpy areas in the powder bod y . The
Harry's Cosmeticology
764
0 00
'-
- 0. 0 00
0
00000
000000
0.•—
uE
a)
C >ULèU>U
s o>>>
uL)c.)L
0
L
a)
a)
•0
0
0 a)
0 0
0
a)
E
I-
a)
0 —
E
ri I- ta)
0 C o.0
a)
'I 0.
.0
The Manufacture of Cosmetics 765
separation of large agglomerates which takes place subsequently in the mill
normally assures the completion of the wetting process provided only that the
binder is correctly chosen. Should the binder still appear to be unevenly
distributed after the passage of the powder through the mill, the product can
often be rescued by passing it through as fine a mesh sieve as possible.
Pearl agents, especially the titanium-coaled micas, present a special problem.
Many of these brittle materials, which depend on their size to achieve the
desired effect, are prone to disintegration in the hammer mill. For this reason,
the usuall y have to be mixed into the bulk after its passage through the mill,
nccsstating an additional mixing operation. The pearl may he added in to the
bulk in the same device used to perform the preliminary coarse mixing and it
ma y sometimes be necessary to pass the bulk finally through a sieve to break up
agglomerates of pearl and to ensure its even distribution.
Batch Colour Correction
It is not unusual for the bulk powder product, even though it has been correctly
processed, to require colour correction in order to obtain a satisfactory match to
the standard. Since any addition of pigment or talc needs to be extended, a
passage through the mill is necessary. A common procedure is as follows. After
the preliminary coarse mix has been completed, a small amount of the hulk
(usually about 5 kg), which is assumed to he representative of the whole, is
passed through the mill. This is examined in the laboratory, and if necessary, a
pigment addition specified. This correction is added to the 5 kg of milled
product, mixed in roughly by hand and the 5 kg is re-milled. The twice-milled
sample is returned to the remainder of the bulk and re-mixed in the original
mixer. A further 5 kg is then removed and the process is repeated until a match
is obtained.
There is a number of minor variations to this procedure which are adopted to
Suit individual companies; the most important of these is the use of pigments
previously extended on talc and stored as such. This has the merit of speeding up
the correction process.
When pearl agents are part of the formulation, unless an un-pcarlizcd
standard is provided, the pearl must be added in the correct proportion to the
laboratory Sample before colour can he assessed. Pearl is only added to the bulk
in the last stage of the manufacturing procedure.
Alternatives to the Hammer Mill
The drawbacks to the hammer mill as used for powder extension have led to a
search for other machines which can fulfil this function more satisfactoril y . The
ideal equipment would probably have the following properties:
(ii It would be capable of breaking up weak particles in the size range 50-05
Am without damaging talc or mica particles of similar diameter
(ii) It would be a low energ y device, consuming little power itself without
heating the powder mixture excessively.
(iii) It would be a batch processing device capable of mixing and extending in
one operation.
(iv) It would be rapid: proces'ung times of less than 10 minutes would be
acceptable.
Harry's Cosmeticology
766
(v) It would not cause excessive aeration of the powder (since this causes
further processing problems in later processing).
(vi) It would be easy to clean.
(vii) Its efficiency would not vary with the cohesiveness of the powder; it would
not be affected by poor flow characteristics.
(viii) It would be quiet and clean in operation.
Other comminution devices have been shown to produce extension, particularly
pin mills and fluid energy mills, yet none seems to work as efficiently as the
hammer mill. In recent years, however, the development of high-speed powder
mixers which are also capable of producing some degree of extension has
brought the industry closer to the ideal. Two types in particular arc worthy of
mention. The first of these is best described as the vertical vortex mixer. The
powder mixture is placed in a vertical, cylindrical chamber and is then
accelerated outwards and upwards into a fluidized vortex motion. The motion
may be produced by compressed air blasted sequentially from a series of nozzles
contained in a lower cone-shaped section; alternatively, a propeller-shaped tool
of 'poor aerodynamic' design may be used which rotates rapidly in the dished
base of the mixing bowl. Mixing and dispersion occur at the point of conversion
of the powder particles (in the upper point of the mixing bowl) by particle—
particle collisions. The second type of high-speed mixer is often referred to as a
'plough—shear' device, because of the unusual shape of the mixing paddles which
rotate on an axial shaft in a cylindrical horizontal mixing chamber. These
paddles cause the powder from all parts of the chamber to be thrown about in
such a way that it all passes rapidly through a zone occupied by a series of rapidly
revolving blades on a separate shaft, referred to as a 'chopper'. The chopper is
largely responsible for the powder extension and may be switched on or off
independently of the main axial drive.
Both types of mixer have been used as partial or complete replacement for the
traditional blender—hammer mill combination. The plough—shear type may also
be used for wet-processing.
Apart from the 'dry' powder processing already discussed, the remaining
ptocesses listed in Table 39.1 involve liquids present in sufficiently large
quantities as to impose fluid characteristics on the mixture. Although there are
similarities between the flow of powders and the flow of liquids it is obviously
easier to set .up and sustain flow patterns in the latter. On the whole this makes
the mixing processes easier to perform and a much larger variety of equipment is
conseguently available to choose from.
Even for liquids, however, the science of mixing has not yet been sufficiently
developed to enable the optimum mixer to be designed for a given task from
purely theoretical calculations. Much of the knowledge we have is empirical, and
has been accumulated from trial-and-error practical experience; we have little
detailed knowledge of many of the mixing processes at work.
where D is the diameter of the impeller, N the impeller speed (rpm), p the
density of the mixture and 17 its viscosity.
—50
768 Harry's Cosmericology
A &PPI -.,/
a S b S C S
Figure 39.3 Rate of shear (S) plotted against viscosity () or apparent viscosity
a Newtonian material
b Pseudoplastic material
c Dilatant material
viscosity. The magnitude of the pseudoplastic effect is variable with the identity
of the fluid, although a fall of 25 per cent in viscosity when the rate of shear
doubled is not unusual.
A related type of non-Newtonian behaviour is the property of 'elasticity'. %s
the name implies, elastic fluids have the ability, on being deformed by the acl ion
of shear, to recover their structure rapidly, with a consequent regain of the
energy absorbed duringdeformation. It is probably accurate to regard e Istic
behaviour as an extreme of pseudoplasticity in which the recovery time i' very
short. Certainly it is true that all pseudoplastic fluids exhibit some degree of
elasticity.
Three uther types of theological behaviour are also worth noting, ilthough
they are less frequently encountered in cosmetics processing. Truly 'plastic'
fluids exhibit viscosity versus shear rate curves similar to those of pseudoplastic
materials; but in this case , a certain force must be applied before any shear (r
flow) takes place. 'Dilatant' materials show the opposite effect, viscosity in-
creasing with shear rate (Figure 39.3).
The term 'thixotropic' is often used erroneously to describe pseudoplastic
behaviour. Thixotropic liquids exhibit a fall of viscosity with time at constant
shear rate.
Flow Patterns. Flow patterns in agitated vessels can be resolved into three
principal types: tangential, radial and axial.
In tangential flow, the liquid moves parallel to the direction of the impeller.
Movement of liquid into the surroundings is small and there is little movement
perpendicular to the blades except in eddies near the tips. Tangential flow may
be observed in paddle mixers operating at low speeds or in liquids of sufficient
viscosity to prevent centrifugl flow from being developed (Figure 39.4).
During radial flow, the liquid is discharged outwards from, the impeller by
centrifugal force. If the moving liquid strikes the wall of the vessel, it splits into
two sections, circulating back towards the impeller where it is re-entrained.
Further turbulence and mixing are induced by the splitting of the flow at the
The Manufacture of Cosmetics 771
Side view
\ ____-y
Top view
Figure 39.4 Tangential flow
wall. There is usually some element of radial flow in agitated vessels, but
flat-blade turbines produce flow patterns that are primarily radial (Figure 39.5a).
Axial flow, as the name implies, takes place parallel to the axis of rotation.
Usually the impeller or propeller blades are pitched so that liquid is discharged
axially—the direction of flow may be from top to bottom of the vessel or
vice versa (Figure 39.5b).
Generally speaking, the most difficult flow pattern to maintain is one of axial
flow.
I I i I
Side view
LLJ1k\ Side view
Top view b Top view
Figure 39.5 a Radial flow b Axial flow
C d
Figure 39.6 Various designs of turbine impeller
The Manufacture of Cosmetics 773
Propeller mixers are restricted to use with low viscosity fluids. They have
pitched blades of which the blade angle varies along the length from centre to
tip. Flow patterns developed by propeller mixers have a high axial component
and the rate of circulation is high. They are usually of relatively small diameter,
typically three-bladed, and are used at speeds between 450-2500 rpm. Such
stirrers are used extensively in the cosmetics industry for simple blending
opejations but are not suitable for the suspension of particles which settle
rapidly or for the dissolution of sparingly soluble heavier materials.
Many portable mixers are of the propeller type. If the mixer is mounted
centrally in the mixing tank (Figure 39.7a), because of the entrainment of liquid
above the impeller, the surface becomes depressed and a vortex is formed.
Generally, vortices are to be avoided because of the low order of turbulence and
the air-entrapment which they cause. When they are mounted eccentrically,
however (Figure 39,7b), turbulence is increased and vortices avoided.
Influence of Vessel Shape
In general, the absence of smoothness and rounded corners in mixing vessels
contributes to turbulence and would be expected to improve mixing times. An
extreme case is the introduction of baffles around the circumference of vessels of
Cylindrical cross-section (Figure 39.7c). It is easy to visualize how the introduc-
tion of baffles could interfere with the flow patterns generated by tangential and
radial flow impellers and cause the suppression of vortex formation in centrally
mounted mixers. Experience shows that baffling improves axial flow and
increases turbulence.
It will also be evident that the ratio of tank dimensions can be an important
factor in determining the efficiency of any mixing process. For example, it is
sensible to perform simple blending, such as alcohol–water, in tall cylindrical
vessels of small cross-section using a propeller (with high axial flow). In the
production of anemulsion of medium viscosity using a radial flow turbine, it
• would evidently be desirable to keep the height of the vessel as small as possible
for a given capacity.
U .
J. Utl
c/ d
_
folding of a volume of the mixture and the physical replacement of it into another
part of the mixer where it is Cut and folded again. An example of this distributive
mechanism is illustrated in Figure 39.9, in which for clarity a volume of a mixture
has been isolated and divided into six equal segments, one of which consists of a
black minor component .5 The cube is compressed to one quarter of its initial
height, cut and reassembled as shown. Redistribution of the minor component
has been achieved which, were the process to be repeated often enough, would
eventually achieve the desired level of homogeneity.
A more recent innovation is the so-called static mixer, of which several
designs are now commercially available. Static mixers are essentially in-line
mixing devices in which mixtures flowing through a pipe are cut and folded by a
series of helical elements in a circular tube (Figure 39.10). These elements
(which do not move—hence the name 'static') turn the flowing mixture through
an angle of 1800. Since alternative elements have opposite pitch and are
displaced 90° to each other, this causes the bulk flow to reverse direction at each
junction, and thus the leading edge of each element becomes a cutting device,
splitting and re-folding the mixture in on itself.
Finally, mention must be made of extruders, in which a helical screw forces
the bulk mixture to flow down a tube. Here, the pressure generated can be
enormous, as in soap-plodding, and such energy can cause materials with the
viscosity of toilet soap to undergo laminar flow. The actual flow pattern
produced is complex, being a combination of prssure and drag flow within the
tube.6
viscosity of the bulk, the method of producing flow and the volume within the
mixture under consideration. For certain applications, however, it is desirable to
generate a very intense degree of shear stress in the mixture, and for this
purpose specialized equipment is available. The uses to which such machines are
put in Cosmetics processing include the breaking up of pigment agglomerates and
their dispersion in liquids, the rapid fracture and dispersion of gelling agents (for
example, bentones, cellulose derivatives and alginates) and the size reduction of
internal phase droplets in emulsion products.
The basic principle on which high shear mixers work is the forcing of the
mixture through a very narrow gap at the highest possible velocity. This may be
illustrated with reference to a widely used device in which the mixture is
entrained by a high-speed rotor, moving in very close proximity to an enveloping
stator, which may or may not contain perforations through which the mixture is
forced. -
Figure 39.11 is a diagrammatic representation of one of the blades of the
turbine separated by a very small gap, h (a few thousandths of an inch), from the
stator. If the velocity of the rotor is v, then thp shear rate, y, is given by
(4)
(y has a typical value of 100-500 s'). It follows from the basic definition of
viscosity given in equation (2) that for a Newtonian liquid having a viscosity r,
hr
(5)
or
71V
(6)
where r is the shear stress. Thus it may be seen that shear stress required to
disperse agglomerates or to reduce droplet size may be increased with increasing
viscosity of the mixture, increasing velocity of the rotor and decreasing clearance
between rotor and stator. For non-Newtonian liquids for which the apparent
viscosity, 17a,p, is given by
7lapp =. (7)
r 17.pp (8)
= h)
This simple treatment does not give the complete picture. Shear stress can be
increased still further by the perforation of the stator, and the flow pattern over
and through the shearing head may be drastically altered by the design of the
assembly enclosing it. Usually, an interchangeable choice of stator designs is
available with the mixer, allowing the most appropriate type to be used for a
particular task. With some designs, it is possible to make use of the considerable
pumping capacity afforded by the high-speud rotor.
Generally, a high shear rotor—stator mixer may be used either for batch
processing in a mixing tank, or as an in-line device when it is encased in a
suitable all-enveloping chamber. Used as a batch mixer, it-is capable of
generating considerable turbulence because of the great velocity with which fluid
is pumped Out of the mixing head. As with other devices, however, this high
energy is increasingly converted into heat with increasing viscosity of the
mixture. A serious disadvantage for certain processes is the tendency of the
mixer to cause aeration when used in the top-entry mode. For this reason, such
devices are often incorporated into the bottom of processing vessels.
The rotor—stator is generally more efficient when it is used as an in-line mixer,
particularly as the mixing time required to ensure that every part of the mixture
has passed through the mixing head is generally reduced (especially for higher
viscosity mixtures). Another disadvantage is that the outer tips of the rotor may
wear out rapidly, leading to increased clearance and decreased efficiency. Since
no adjustment is possible, these must be replaced at considerable cost.
Another high shear rotor—stator device in common use is the colloid or stone
mill. Such equipment is commonly thought of as a comminution device; this
serves to illustrate the fineness of the dividing line between mixing and
comminution with high shear equipment. While it is true that colloid mills may
be used for the comminution of very soft materials in a slurry, they find
application in the cosmetics industry for the dispersion of pigments and the size
reduction of internal phase droplets in emulsions. In principle, the colloid mill
consists of a rapidly rotating conical member (which may be toothed or grooved)
and a similarly coned stator into which the former fits. The fluid mixture is
forced through the small clearance between rotor and stator (0 . 5-0 . 05 mm) as
before. Figure 39.12 illustrates the design of colloid mills in greater detail.
Colloid mills are used exclusively as an 'in-line' or continuous device. They
may be water-cooled and can be adjusted as the moving parts wear down.
The Manufacture of Cosmetics 779
] f:
For more viscous products, an alternative device is the triple roll mill
(Figure 39.13). The device consists of three steel rollers which rotate in the
directions indicated in the diagram. Each roller is water-cooled and machined to
great accuracy so that the gaps between each pair of rollers can be set very fine.
The product is applied at the top of roller A, passes between A and B and round
the underside of B between rollers B and C. As it traverses each gap, the
product is subjected to enormous compression forces which are particularly
effective in breaking down pigment agglomerates. Strictly speaking, therefore,
the triple roll mill is not a high shear device but a compression device. It is
included here, however, as a real alternative to the rotor–stator mixers in
bringing about effective pigment dispersion in liquids—particularly in viscous
liquids.
Perhaps the highest shear stress of all is generated by a valve homogenizer,
which is still extensively used in the production of emulsions with very fine
internal phase droplets. A.valve homogenizer (Figure 39.14) is simply a high-
pressure pump which forces the product through a small orifice at pressures of
up to 350 atmospheres (354 bar).
0 00 G
o%o00%o%
S 0 000600o
00 0
0 00
0%
0000
0%oo
0 0
0 0 00 0
0000 00
000 00
0 00 000
00 0
00
000
Ooo 00
00
00 00 0
00 0 0 0
00000
0000
'00 0000
0
)1'•
Figure 39.16 Ball mill showing ball pattern in rotating drum
G Grinding medium
L Lifter elements
S, Slurry
782 Harry's Cosmeticology
Small ball mills and sand mills are used extensively in the dispersion of
pigments into liquids (as; for example, in the production of castor oil lipstick
pastes) and for the dispersion of bentonc into nail varnish media.. Although they
are very effective, their chief disadvantage is the extremely protracted cleaning
time required when changing from one colour to another. For this reason many
users prefer to keep separate sets of grinding media for each different colour
they wish to produce.
The basic mechanism by which mills of this kind produce their effect is
attrition between grinding elements. Very little shear is developed.
(9)
D5N3p
where P is the power imparted to the fluid of density p, and D is the diameter of
the impeller which has a rotational speed of N. For many pieces of mixing
equipment, P is measurable from electrical consumption data (provided that
friction losses in gearboxes and bearings are ignored) and hence P. can be
calculated.
It has already been pointed out in equation (1) that another dimensionless term,
the Reynolds number (Re), is useful in describing the onset of turbulence. The
relationship between P. and Re has the general form indicated in Figure 39.18.
The value of graphs such as this lies in the insight they give into the
interdependence of flow and mixing behaviour and the-design and operating
characteristics of the mixer. This should help the process engineer to choose
not only the best equipment for a particular task but also the best conditions
under which to operate it.
Figure 39.18, for example, shows that when flow in the mixing vessel is
non-turbulent, the power which the agitator applies to the mixture falls rapidly
as the agitator speed increases until, at a Reynolds number of about 200, it
flattens out. If the vessel is unbaffled (curve 5), the power input then decreases
again as vortex formation takes place. Here, the maximum power input that can
be achieved occurs when the vortex just reaches the turbine.
The other curves show the effect of four different baffle widths. By the time
the increased agitator speed has taken Re to 10 4 , turbulence has fully developed.
Notice that at this point the power input becomes independent of Reynolds
number and dependent upon the extent of baffle. The task of the production
chemist is to choose the Reynolds number and degree of baffle which will
achieve the desired end result with the minimum power input.
Mixing Time. Another important measurable parameter is the mixing time, tm.
This is- the time taken to achieve the desired degree of homogeneity in the mixer.
—51
784 Harry's Cosmeticology
102
4
P. 10
3
2
Re
Figure 39.18 Relationship between power number, P0 . and Reynolds number, Re
1,2,3,4 represent increasing degrees of baffle
5 represents an unbaffled tank
There are many methods by which this characteristic may be measured but
perhaps the most obvious is the time taken for a soluble dye to become
uniformly dispersed throughout the mixing vessel (as, for example, in the
manufacture of a coloured shampoo). The relationship between mixing time, tm,
and the degree of uniformity can clearly be shown if some index of mixing can be
established.
A simple example of this would be the ratio of colour intensity between the
top and bottom of the mixer contents at intervals after dye is added to the top (so
that uniformity is achieved as the mixing index, M, approaches unity). This
comparatively simple experiment should give rise to a curve similar to that
shown in Figure 39.19. Since the approach of M to unity is asymptotic, tm is
difficult to measure accurately unless a colorimeter or other optical colour
measuring device is available.
Once tm has been established, however, more useful insight into the para-
meters controlling mixing rates may be gleaned from relationships such as that
illustrated in Figure 39.20.811 In Figure 39.20a (which relates to a viscous liquid
in which turbulence is not established), tm, the mixing time, has been replaced by
10
0 tm
time, t
Figure 39.19 Mixing index, M, plotted against time to give mixing time, t.
The Manufacture of Cosmetics 785
0 0
Nt,,, Ntrn
a b
Figure 39.20 Mixing index, M, plotted against number of revolutions, Ni,,,
a Newtonian fluids b Non-Newtonian fluids
the product of rotational speed N and tm' that is, by the number of revolutions of
the impeller.
It is interesting to note that for Newtonian fluids precisely the same plot is
produced whatever the viscosity of the medium and spe.d of impeller. In other
words, only the number of impeller revolutions determines the change in mix-
ing index. This is not true of non-Newtonian fluids—plots A,B and C in
Figure 39.20b represent liquids showing increasng divergence from Newtonian
behaviour. This illustrates the difficulties, already mentioned, of mixing
non-Newtonian media, in which flow is damped Out rapidly by regons of
highs viscosity away from the vicinity of the impeller blade.
The mixing time may be converted into another dimensionless group, for
example t,,,N D3 /v, where v = volume. The relationship between mixing time
and the development of turbulence can then be elucidated. Figure 39.21 shows
the general shape of curves most commonly obtained by plotting dimensionless
>
0
z
C
10 102 103 10
Re
Figure 39.21 Dimensionless mixing time versus Reynolds number
Harry's Cosmeticology
mixing time against Reynolds number.' , " Mixing time is seen to decrease
rapidly as Re increases through the laminar region. Once full turbulence has
been established, however, further increase of impeller speed has little effect.
Pumping Capacity and Velocity Head. Perhaps the most powerful concept to
arise from the analytical approach to mixing concerns the way in which the
power provided by each type of impeller is actually transmitted to the fluid. This
relationship can be expressed generally as
PcxQH (10)
where Q is the pumping capacity of the impeller (the volume of fluid displaced
directly by the impeller in litres per minute) and H is the velocity head—this i
related to the shear rate experienced by fluid leaving the impeller. A large
slow-moving impeller might produce, for example, a large pumping capacity and
a low-velocity, head, while a small impeller operating at high speed might
produce a lower volume of fluid pumping but at a much higher velocity head.
Some cosmetics production processes—perhaps the majority—require a high
pumping capacity; others require a high shear rate or velocity head. It is
therefore useful to know the parameters that affect both these functions and how
they interrelate.
For simple blending operations (the manufacture of shampoos or colognes,
for example) pumping capacity is often of the greatest significance. Under
conditions of laminar flow, the number of complete circulations of the bulk
(turnover') required to bring about homogeneity is approximately three. For a
turbine operating in turbulent conditions, this is reduced to about 15. However,
given that only a fixed amount of power is available from the motor, it is not
likely that a relatively small turbine will have sufficient pumping capacity to push
a fairly viscous product around by even this amount.
Not surprisingly, the factor which determines whether power is used as
pumping capacity or velocity head is the ratio of impeller to tank diameter
(D/fl. Experiments have established the following relationship over a wide
range of condftiàns
q) I
(H
where the first term is the ratio of Q to H at a constant power. As shown in
Figure 39.22, however, there is little point in a bIT ratio of beyond 0.6. 12 The
total flow includes flow generated by entrainment of quiescent liquid in the
direct flow from the impeller head and may be several times greater than Q.
The picture is completed by the relationship between power consumption and
DI ratio for equal process results in a given vessel (Figure 39.23) .. The use of a
larger D/T ratio lowers the power required to achieve the same end result. At
the same time, this implies the use .of lower impeller speed and this inevitably
means that the torque required to drive the mixer increases dramatically.' 2 The
amount of torque that a given mixer is able to accommodate depends largely on
its construction. It therefore becomes a question of economics whether to invest
in a more substantial (and therefore expensive) mixer in order to reduce the
power consumption needed to achieve a given mixture quality.
Constant power
08
0-6
0-4
0-2
2-0
1-5
1-0
09
0-B
0-7
0-6
05
041 1 1 i I I I
Scale-up
It is not unusual for a product that has been developed and made successfully in
the laboratory (using laboratory stirrers and beakers) to exhibit quite different
characteristics when made on a production scale. This is not due to the
incompetence of the manufacturing department but represents the difference in
conditions experienced by the product because of the change of scale. Even if a
pilot plant is available of intermediate scale, there is no guarantee that severe
problems will not be encountered during manufacture of the first full-scale
batch. The study of scale-up is therefore of fundamental importance to efficient
cosmetics production. Unfortunately, it is also an extremely complex subject
since the variables that determine the distribution of forces within the processing
vessel vary considerably as the scale changes. 12
The quantities that are most often considered in scale-up are total power,
power per unit volume of liquid, impeller speed, impeller diameter, the flow
from the impeller, flow per unit volume, peripheral speed, D/T ratio and
Reynolds number. For a ten-fold increase in scale in geometrically similar
vessels, it is impossible to keep all of these constant. For example, if power per
unit volume is held constant during scale-up from 200 litres to 2 tonnes, then
every other parameter varies.
The task of the process chemist is therefore to understand which characteris-
tics are controlling the process, so that he knows which parameters must be kept
constant and which may be allowed to vary. For example, the manufacture of an
emulsion requires turbulence (Reynolds number) to be held constant in order to
achieve consistent droplet size of the internal phase. At the same time, the flow
must be great enough to ensure that all the constituents pass through the
turbulent regions.
In general, it may be said that there is not always sufficient awareness on the
part of the development chemist of the problems arising from scale-up. There is
no point in producing a superb new product in the laboratory which it is
impossible to mike in the factory. A pilot plant experiment can be a valuable
tool in choosing the equipment and processes needed for full scale production
but not if the study is made in a superficial way. This is an area where a
fundamental understanding of mixing processes and a willingness to experiment
will prove most fruitful.
SOLID—LIQUID MIXING
crystal into the liquid involves the complete replacement of the air–solid surface
with a liquid–solid surface. This may be considered to be a three-stage process of
adhesion, immersion and spreading (Figure 39.24). Immersion is complete when
all the air has been displaced and the surface of the crystal has been completely
wetted by the liquid. This process is aided by a low surface tension and low
contact angle between the liquid and solid.
Not all powders used in cosmetics, however, have such favourable size and
surface characteristics. The majority are of extremely small particle size and, as
has already been noted, highly agglomerated. Each agglomerate will have a
complex structure with an uneven surface and will be perforated by cavities of
irregular shape. The complete wetting of such structures, involving the penetra-
tion of liquid into all the crevices and cavities together with the expulsion of air,
is very much more difficult. It should be noted, for example, that penetration
into cavities requires a low contact angle but a high surface tension—in conflict
with conditions for easy wetting.
Even more complex is the immersion of powders which swell in the liquid to
form dispersions of colloidal size, since the particles on the outside of each
agglomerated mass tend to swell and adhere to each other, slowing down
penetration of the liquid to the still-dry core.
Powders that are small enough in particle size to form agglomerates are by far
the most commonly used in cosmetics processing. In the dry state they entrap an
enormous quantity of air (a bag of cosmetic-grade titanium dioxide, for
example, contains only 25 per cent of powder together with 75 per Cent of air).
The niajority of this air must be expelled if a smooth uniform mixture is to be
obtained.
Immersion is only the first stage in the production of cosmetic quality
dispersion. Even if the agglomerates were evenly distributed, the larger of them
would give rise to 'grittiness'. Further, for pigments maximum colour can only
be developed when these agglomerates are broken up and the maximum,
possible surface area of pigment is exposed. Disagglomeration is therefore the
next step in the production process.
The forces holding these agglomerates together are precisely the same as
those described earlier in the chapter in the section on powder mixing. The
obvious difference, of course, is that these agglomerates are situated in a fluid
medium, the physicochemical characteristics of which may enter into the
reckoning of bond strength, ease of separation and likelihood of reagglomera-
tion. Consequently, the theoretical treatment of particle–particle interaction in
Air A
z:
Liquid
d\\
Figure 39.24 Immersion of a solid in a liquid
A Adhesion B Immersion C Spreading
790 Harry's Cosmeticology
liquid media is even more complex than for dry solids—although this has been
extensively studied and written about.
In cosmetic processes, the disagglomeration of solid particles in liquid media
can be brought about by a variety of machines. In lipstick processing, for
example, pigments are 'ground into' castor oil by preparing a coarse mixture
Which is then passed over a triple roll mill, or further ground in a colloid mill,
ball mill or sand mill. These machines are used specifically because they will deal
effectively with media of lipstick-paste viscosity.
For less viscous media (for example, the dispersion of pigments into the
aqueous phase of an emulsion), a high shear device of the rotor–stator type is
frequently used. In this case, processing time can be shortened by ensuring that
the whole contents of the vessel are brought into the .catchment area of the
shearing head by secondary stirring. As with all disagglomeration, shear stress is
largely responsible for the partial disintegration of the agglomerate.
For soluble powders, the enormous increase in the solid–liquid interface
brought about by immersion and disagglomeration will ensure that the actual
process of dissolution can proceed at the maximum possible rate. For insoluble
powders, however, there remains the problem of maintaining a good stable
dispersion.
Disagglomeration is usually a reversible phenomenon and it can usually be
assumed that the opposite process—'flocculation'—will be simultaneously
taking place.
It was noted when considering powder–powder dispersion that stabilization
could be achieved by the introduction of particles of larger size to which the
disintegrated agglomerates could adhere. In some cases this can be
applied to solid–liquid systems—for example, by pre-extending pigments onto
talc before adding them to a liquid foundation base—but in many instances all
the solid particles are of too small a size for this to be done. Under these
circumstances rules similar to those used in emulsion technology can be applied.
Thus the rate of flocculation can be slowed by some or all of the following
means:
(i) The usd of surface-active agents (sometimes as polymer coating of the
powdered solid) to inhibit flocculation by steric hindrance.
(ii) The manipulation of electrostatic charges on the surfaces of the powder
particles.
(iii) The manipulation of the viscosity of the dispersion.
Surface-active agents have a part to play at two stages in the process of
manufacturing a stable dispersion. It has already been seen that the lowering of
the solid–liquid contact angle speeds up the wetting process. In practice, the best
results are often achieved not with a surfactant, which measurably lowers the
surface tension of the liquid, but with what is sometimes described as a 'surface
activator' which reduces the interfacial tension between solid and liquid. These
surface activators (which are also described as 'dispersants' or 'wetting agents')
can, if correctly chosen, cause an immediate improvement in the quality of the
dispersion which is made manifest by a sudden increase in the colour intensity.
The rules for choosing a wetting agent or dispersant are similar to those used
for surfactants in emulsions: part of the molecule must have affinity for the liquid
The Manufacture of Cosmetics 791
medium and part for the solid. Clearly, if those parts of the molecule which have
affinity for the liquid are large and present in sufficient numbers, by spreading
• out around each solid particle they can form a physical barrier that prevents
particles from coming close enough together to reagglomerate. In this way an
iron oxide dispersion in castor oil, for example, could be stabilized by the
addition of a wetting agent having a hydrophilic end which would adhere to the
oxide surface, and a long unsaturated fatty acid end which would 'be able to
extend Out into the castor oil medium surrounding each particle.
Where the liquid medium has a sufficiently high dielectric constant, other
types of dispersion can be stabilized by surface activators having a residual
electrostatic charge associated with them, thus preventing the flocculation of
particles by the mutual repulsion of like charges.
A significant difference exists, however, between emulsions (where the two
phases must perforce have very different chemical affinities) and dispersions dl
solids in which, for example, a hydrophilic surface can be dispersed in water.
• When this happens, this matching of affinities is an advantage. This has led to the
processing of powders in order to coat their surfaces with a chemical (usually
polymeric) of suitable characteristics to allow easy wetting and dispersion.
Nowhere is this more clearly illustrated than in the case of water-wettable and
oil-wettable grades of titanium dioxide. In this instance, the same grade of
titanium dioxide can be coated with different resins to modify the surface in such
a way as to make it wettable by either water or oil.
In aqueous media, the possibility exists for the particle surfaces to become
electrostatically charged. This is an important consideration in the discussion of
flocculation. For example, the effect of pH on the quality of pigment dispersions
in emulsion-based products is often overlooked until it causes a problem by
giving rise to an unexpected colour change. For those dispersions, particularly
oxide surfaces, in which electrostatic repulsion is part of the stabilization
mechanism, flocculation can occur rapidly as the isoelectric point is approached,
if the pH is allowed to vary during manufacture.
Naturally, since particles have to move towards each other in order to
flocculate, the viscosity of the medium through which they must move has apart
to play in the rate of flocculation. It is necessary, however, to distinguish
between the viscosity of the total dispersion which, as will be shown, is
influenced by the solids content, and the intrinsic viscosity of the liquid medium
itself; it is the latter which predominantly influences the rate of flocculation in
low solids-content dispersions. The addition of thixotropic gums to nail lacquers
and of colloidal thickeners to the aqueous phase of emulsions serves to slow
down flocculation without materially influencing the basic flocculation process
itself. For this reason the reheating of liquid pigmented foundation products
sometimes results in unexpected changes of hue which are often erroneously
ascribed to phase-inversion. The truth is that the rate of flocculation of an
intrinsically unstable dispersion has been speeded up because of a drop in
viscosity caused by the heating process.
Notwithstanding the viscosity of the liquid phase, it is generally true that the
viscosity of the dispersion increases with the solids content, as does the difficulty
of maintaining a good dispersion. For a given solids content, viscosity decreases
with the particle size of the solid phase, and thus disagglomeration is usually
792 Harry's Cosmeticology
LIQUID–LIQUID MIXING
Miscible Liquids
The mixing of miscible liquids (blending) represents perhaps the simplest mixing
operation in cosmetics manufacture. Several examples have already been cited
and no further elaboration is needed except to reiterate that it is important to
choose the mixing apparatus best suited to the viscosities of various components
in order to carry out the operation efficiently.
Immiscible Liquids
Practically the only representatives of this category of mixing operation are
emulsions. The theory of emulsions is covered fully in Chapter 38; briefly, all
cosmetic emulsions consist of two major immiscible liquids, one dispersed as fine
droplets in the other and separated by a layer of surface-active agent at each
liquid–liquid boundary. (This is a simple view. In practice the two main phases
are not always liquid at room temperature, and the total number of composite
phases may be greater than thre—this in no way invalidates a general
discussion of the process available for the manufacture of emulsions.)
794 Harry's Cosmeticology
Orientation of Phases. In any emulsion the orientation of the phases (that is,
whether the oil or the water phase is continuous) is determined principally by the
choice of emulsifier and the volume ratio of oil to water. Usually, however, there
is a range of volume ratio over which either phase may be dispersed, depending
upon the method of manufacture. If a quiescent mixture of two phases coexisting
as two simple layers (one upon the other) is agitated, the phase into which the
agitator is placed is most likely to form the continuous phase in the resulting
emulsion. In other words, drops are drawn into the phase in which the impeller
is placed. If initially only one phase is present in the mixing vessel containing the
impeller, an added second phase will inevitably form the disperse or discon-
tinuous phase. If, however, the continued addition of the second phase
combined with the choice of emulsifier eventually leads to a volume ratio at
which the system is more stable with the added phase being continuous, then the
emulsion will spontaneously invert to achieve this end result. When inversion
takes place, it is very often accompanied by a change in droplet size. Where this
The Manufacture of Cosmetics 795
C0
0
-J
Log N
Figure 39.25 Relationship between drop size, ci,, and stirrer speed, N
change is a decrease, then inversion leads to a more stable emulsion and gives
rise to a valuable method of manufacture.
Addition of Surfactant. In a batch manufacturing process for emulsions, there
are four possible methods of adding the emulsifier. The first of these involves
dissolving (or dispersing) the emulsifying agent in water, to which the oil is
added. An oil-in-water emulsion is initially produced but inversion to water-in-
oil may take place if more oil is needed.
Alternatively, the emulsifier may be added to the oil phase; the mixture may
then be added directly to water to form an oil-in-water emulsion or water may be
added to the mixture to form a water-in-oil emulsion. Many emulsions, on the
other hand, are stabilized by soaps which are formed at the interface between
the two phases. In this case, the fatty acid end of the-soap is dissolved in the oil
and the alkaline component is dissolved in the water. The two phases can be
brought together in any order.
Finally, a less used method is one in which water and oil are added alternately
to the emulsifying Vent. Usually, the improvement in product quality obtained
by the use of this method does not warrant the complication it causes in the
manufacturing procedure.
Batch Processing Equipment. It will be evident from the discussion so far that
there are at least two important elements of emulsion processing, namely shear
(for the emulsification and particle size reduction process) and flow (in order to'
796 Harry's Cosmeticology
bring the whole contents of the vessel through the region of high shear). Flow is
also important in the heating and cooling of the emulsion. Most emulsion
processing vessels are equipped with a jacket through which steam or hot water
can be circulated to heat the contents and cold water circulated to cool them.
Evidently then, to be effective, the mixing mechanism must be able to provide
adequate flow to and from the vessel walls.
For these reasons, most emulsion batch processing vessels contain a high shear
turbine or rotor—stator device (typically bottom or side entry rather thau top
entry, to decrease the likelihood of air entrapment) and a high flow, low shear
mixing device which may be driven by a separate motor. This high flow device is
of variable design, the most popular being a gate stirrer in which the arms are
inclined at about 450 to the horizontal so as to give an element of axial flow. In
more complicated designs,. a central shalt carries more blades which sweep the
area between the first set, the sets of blades rotating in opposite directions.
Whatever the design, the frame holding the outer blades normally carries
spring-loaded plastic scraper blades to prevent the build-up of product on the
inner vessel wall, which would interfere with efficient heat exchange across the
surface (Figure 39.26).
The main motor may be driven by electricity or by air (up to about 100 psi or 9
bar). The advantages of air-driven motors are that they are infinitely variable in
speed, torque-sensitive (and therefore less likely to become damaged when
subjected to sudden loads), they do not constitute a hazard in the processing of
low flash-point materials and generally they require less maintenance. Electric
motors can be built to match some of these advantages (with slipping clutches
and flameproofing), but only at considerable expense.
This traditional gate-type impeller system suffers from the grave disadvantage
of limited axial flow. This is not noticeable in smaller vessels (below 600 litres
capacity) but becomes a major problem in large tanks. One approach to the
problem is to provide top-to-bottom transfer of the contents by means of a pump
and an external pipe. A more satisfactory arrangement for the manufacture of
emulsions, of medium and low viscosity is to replace the gate stirrer by one or
more axial flow impellers mounted centrally on a single central shaft. Although
it becomes more difficult to provide wall-scrapers, the excellent flow aroun&the
vessel walls makes scraping less necessary.
The problem of ensuring that all the product passes through the reo p of high
shear has led to the idea of passing the batch through an external circuit
containing an in-line homogenizer; this may be a rotor—stator device, colloid
mill or valve homogenizer.
Outer
blade
team/water
cket
Scraper
blades
Inner blade
shear
ce
be prepared while the first is being used. In this way, a Continuous supply of
each phase is assured by the turning of a valve. In reality, continuous plants tend
to be slightly more complex than is illustrated in the diagrams with the inclusion
of take-off points for sampling and other sophisticated features. Nevertheless,
continuous manufacture is a very practical and, for some applications, extremely
economical method of processing emulsions.
Emulsion Temperature. The primary reason for raising the temperature of the
phases during emulsion manufacture is to ensure that both are in the liquid state.
In particular, the oil phase may contain fats and waxes that are solid at room
temperature; there is very little point in raising the temperature of the oil phase
much above that at which these liquefy. Excessive heating of the phases during
manufacture prolongs the manufacturing time and wastes energy.
If the water phase is liquid at room temperature, it is customarily heated to
approximately 5°C above the temperature chosen for the oil phase (so as not to
cause the sudden solidification of the latter on blending). There is, however, an
interesting alternative—namely, emulsification between hot oil phase and cold
water phase. The plant for this procedure is illustrated in Figure 39.29, which
shows that mixing of the phases and homogenization take place simultaneously.
The obvious advantage of such a method is the saving of time and energy in not
having to heat the aqueous phase.
REFERENCES
Li
Figure 39.28 Continuous-processing emulsion plant for Tong-run products
Aerosols
Introduction
Despite the fact that there had been a series of patents relating to the packaging
of products under pressure going back as far as 1899 , 1 it was not until the 1940s
that the principle was applied for the presentation of consumer goods. In 1943,
as a result of research on pesrcontrol, Goodhue and Sullivan were granted a US
patent 2 for a portable aerosol dispenser, in the form of a heavy gauge metal
container filled with insecticide and dichlorodifluoromethafle having an internal
pressure of approximately 70 psig (483 kPa). These portable aerosols were used
at the time by the United States armed forces.
In 1947, USA statutes relating to pressurized containers were modified to
permit the use of thin-walled containers more suitable for the consumer market.
A rapid development of containers and valves then followed, giving rise to a new
packaging industry. Because the parent pack (the wartime insecticide) dispensed
its product as a true aerosol, all the descendant pressure packs, irrespective of
the physical form of the product dispensed, have been subsequently called
'aerosol'. While for many products the term 'pressurized pack' is more accurate,
the original designation is still widely used.
The new form of packaging enjoyed a spectacular growth rate. Thus in the
USA a market of just over 5 million units in 1947 increased to 2000 million in
1967. World fillings were close to 4000 million in 1968, and increased to 6000
million in 1978.
Over 6500 million were filled in 1980, just over one-third in the USA and
another one-third in Europe. In these two geographical areas, personal prod-
ucts—mainly hair sprays and deodorant/antiperspirant—form just over half
N the market, with household products (insecticides, polishes and air fresheners)
forming the majority of the remainder. For the rest of the world insecticides are
predominant but most future growth is expected in personal care products.
THE AEROSOL
CONTAINERS
Containers are available in tinplate and aluminium, plas ic-coated glass and
plastics. In America seamless drawn steel cans, usually int'rnally lacquered and
fitted with a tinplate. base, can be found.
Tinplate
More than half of the world's aerosols are packed in three-piece tinplate
containers with welded or soldered side seams. They are available in many sizes
from 4 oz (150 ml) to 30 oz (1000 ml). The amc unt of tincoating applied varies
with the component— bod y , cone or dome—and the country of origin,
(2 . 8 g m 2 ) being common for bodies in America and Europe. E28
There are three types of solder generally used for making side seams:
1. The 2/98 tin/lead system (mixed solder or standard solder).
2. The Duocom (antimony/tin) system.
3. The 100 per cent tin system (plain solder).
The advantage of using plain (100 per cent tin) solder is that it does not give rise
to the internal or external corrosion which may occur with mixed solder;
however, because of its inferior mechanical strength it is liable to give rise to
troubles at high internal pressures.
Only containers with welded seams are made now in the United Kingdom5
and welding is gradually replacing soldering in other countries. The process is
based on technology rather than craftsmanship and has therefore potential for
further development. The weld is stronger than the body plate, allowing higher
pressure formulations to be packed. The seam width is only a quarter that on
soldered cans and thus.it is possible t9 get nearly all-round decoration.
Aluminium
Aluminium cans are of two main types:
1. One-piece containers, the monobjocs, available in sizes of up to 36 oz
(1200 ml).
2. Two-piece containers in sizes of up to. 20 oz (750 ml).
Harry's Cosmeticology
802
These also differ in the neck profile, that of the monobloc being in the form of
a rolled bead, while two-piece containers have solid beads which are not as
consistent as rolled beads and are more prone to ridges and dents.
The two-piece cans have an aluminium body and a seamed-on base of either
aluminium or tinplate. The latter type has the advantage of allowing the use of a
magnetic water bath, but this is to some extent offset by a greater likelihood of
corrosion. Both tinplate and aluminium frequently require protection from
corrosion by the contents of the container. This is generally achieved by
lacquering or, in the case of aluminium, by anodizing (see later in this
chapter—Corrosion).
Uncoated Glass
Uncoated glass aerosol containers are attractive, with freedom of shape, but as a
safety measure' only low pressures (between 15 and 20 psig; 103-138 kPa)
should be. used.
Plastic-coated Glass
The use of plastic-coated glass containers is restricted by pressure considera-
tion. Vinson 7 recommended a PVC coating with a thickness of not less than
0. 015 in (038 mm).
Valves for these containers are more expensive, since bottle valves are used
instead of standard one inch valves and they are not produced in large
quantities. The crimping method is different, and involves external crimping or
swaging. The rejection rate of bottle valves is large.. Advantages of glass packs
are that they are non-reactive and cause no corrosion problems, since valves are
not in contact with metal.
Selections of the various types of metal and glass containers available at
present are illustrated in Figures 40.1, 40.2 and 40.3.
Plastic Packs
Plastic containers for aerosols are a more recent addition to this field of
technology. They combine the advantages of safety and freedom from corrosion.
However, they are rather expensive and in some cases interactions may occur
between the perfun'te and the plastic or the plasticizer present. The plastic
materials used include polyacetal and polypropylne.
Safety Considerations
Modem metal containers, especially those without a side seam, can be used with
safely for pressures of up to 100 psig (690 kPa) at 21 CC but glass packs containing
liquefied gas propellants are only suitable for lower pressures. Even then they
may require to have a metal, cardboard or plastic protective cover, which may
be integral with the glass pack.
Aerosols 803
I
rim
Figure 403 Glass aerosol containers
(Courtesy Max Factor Ltd)
VALVES
The most important part of the dispenser is the valve. There are many types of
valve available, the major manufacturers being the Precision Valve Corp.,
Aerosol Researc and Development, the Risdon Manufacturing Co., the
Newman-Green Valve Co., Seaquist, Ethyl Corporation and Coster Tecnologie
Speciali S.p.A.
The details of construction and operation of some of the main types of valve
are outlined below (see Figure 40.4, P. 819).
The valve is the mechanism for discharging the product from the container,
and in aerosol technology the term refers to the entire assembly which is sealed
on to the aerosol pack, including the valve stem, spring and valve housing or
body, the metal cup in which it is mounted, the dip tube (when present)
connecting the valve with the contents of the container, and finally the actuator
button or spray tip through which the product is discharged into the atmosphere.
Figure 40.4 illustrates the intricacies of a typical valve assembly, and shows the
propellant injection paths during the filling stage. The gasket, which may be of
one of many materials (for example, nitrile, neoprene, butyl Or Viton), plays a
vital rote in the operation of the aerosol and maintenance of pack-integrity
throughout its life. It is important, therefore, to apply deliberate care and
attention to the correct choice of gasket, in order to achieve complete compati-
bility with the product and to ensure optimum valve performance. The valve
determines the rate of discharge of the product, the form in which it is
discharged—spray, foam, cream or powder—and, in the case of sprays, the
Aerosols 805
fineness, the spray profile and spray pattern. By using metering valves, it is also
possible to discharge a predetermined amount of product each time the valve is
operated. The choice of valve and its various components is just as important in
determining the success of the product as the selection of the propellant
composition and the choice • of product constituents.
Components
The basic components of the valve mechanism are:
1. The dip tube through which the product is delivered from the bottom of the
container into the housing. It is usually made of polyethylene or polypropy-
lene.
2. The valve housing or body which holds all the valve components together,
and which itself is held in the mounting cup.
3. The valve, which transmits the product from the housing to the actuator
button or spray tip and which acts as an expansion chamber.
4. The stem, a component of the valve, upon which the button is placed.
5. The gasket, which seals the stem orifice when the valve is closed.
6. The spring, which returns the stem to close the valve when the pressure upon
the actuator button is released.
7. The actuator button (spray tip), which operates the valve and which controls
product discharge. It can be made from various plastics, including nylon,
polyethylene, polypropylene and acetal. The orifice size of the button will
vary within a fairly wide range depending on the product to be dispensed.
8. The mounting cup, which is crimped around the housing and mounted on the
container.
9. The cup lining or inner gasket, which is either flowed into, or placed in, the
rim of the valve cup to form a seal when the valve cup is crimped over the
neck of the container.
Operation
The presence of a compressed or liquefied gas in the sealed container gives rise
to a pressure differential between the head space inside the container and the
surrounding atmosphere. When the valve is opened, the product is pushed up
from the base of the container through the dip tube into the valve housing, then
through the open seal, up the valve stem, through the valve stem orifice into the
expansion chamber formed by the valve stem and the actuator passage, and
finally emerges into the atmosphere as a jet, or fine or coarse spray or foam,
depending on the design of the valve and actuator button, as well as on the
propellant system.
Types of Valve
There are many types of valve designed to meet various specialized needs and
purposes. Some of the variations are described below.
806 Harry's. Cosmeticology
Standard Valves
These valves are designed to fulfil the function of controlling the discharge of the
product from the container in a simple on—off manner by breaking or making a
seal within the valve. The rate of discharge is governed by either the orifice at
the seal or that at the final exit, whichever is the smaller, and ths can be selected
according to requirements. There are several variations of mechanism which
come within the description of standard valves; these include vertical or tilt
action buttons, integral or separate movable valve stems, and various designs of
springs and seals.
Foam Valves
The valves used for dispensing foams are essentially standard valves fitted with
wide unobstructed passages on the exit side of the seal orifice. This passage,
which ends in a foam spout, serves as the expansion chamber in which the foam
forms. The type of foam is governed by the composition and proportions of
product and propellant.
Powder Valves
These also are little different from standard valves and the principal requirement
is a smooth flow passage on both sides of the valve seal for the slurry of powder
and propellant, so that there is a minimum possibility of powder deposition to
interfere with the valve operation. Correct choice of powder ingredients and
propellant is essential.
The Risdon Company claimed to have overcome the problem of valve
blockage by powders by means of a 'slide-and-clean' valve in which the internal
metering orifice is in a movable stem, and which cleans itself by passing through
the seal each time the actuator button is released. The stem wall carrying the
orifice is thin so that the powder film in the orifice is disrupted by the pressure
when the valve is opened. A number of other valve companies, notably the
Precision Valve Corporation, successfully introduced long-stroke valves to
prevent powder build-up and blockage of the orifice.
Metering Valves
When it is desired to dispense a fixed amount of product for each actuation of
the button, this is achieved by providing a chamber with sealed entrance and exit
orifices. With the button in the closed position the metering chamber exit is
closed and the entrance open, thus allowing the chamber to fill from the
container. Depression of the button closes the entrance, trapping a metered
volume of product (usually of the order of 50 l) and then opens the exit to
allow the escape of the product for use. -
The vapour phase tap may, however, result in the loss of tuo much propellant
from the head space with the risk of incomplete discharge of the product, or a
changed spray pattern as the Container is emptied. This can be prevented by
ensuring that the tap is of a suitable size and that there is an adequate amount
of propellant present.
Transfer Valves
These valves are used to provide a connection with smooth flow between the
stems of two standard valves with their buttons removed. They are used for
recharging the smaller daily use pack of, for example, 'mother and daughter'
pack hair lacquers, for refilling cigarette lighters with liquid butane and in the
laboratory for pressurizing aerosol cans with propellant from other cans.
Speciality Valves
An innovation in 1968 was Risdon's valve which was designed to work effectively
in any position, from upright to inverted, without loss of vapour gas. 10 The valve
has, in addition to the usual assembly, a double dip tube with a transfer
connection opened and closed by a ball valve to allow the product to flow
between the tubes as the container is tilted. It is claimed that the valve can be
used with all types of propellants and with any actuator.
PROPELLANTS
Chiorofluorocarbon Propellants
The most important chiorofluorocarbon propellants are trichiorofluoromethane,
dichlorodjfluoromethane, dichlorotetrafluoroethane, and their mixtures.
In addition to the 'Freons' of E. I. du Pont de Nemours Co., USA, who were
the first to produce them on the commercial scale, trade names under which
these products are sold include:
Algofrene (Montecatini Societa, Italy)
Arcton (Imperial Chemical Industries Limited, UK)
Forane (Elektrochimie Ugine, France)
Frigen (Hoechst Chemicals, Germany)
Isceon (ISC Chemicals Limited, UK)
An international code is used for the various chlorofluorocarbon compounds,
and is interpreted as follows:
(i) The right-hand number represents the number of fluorine atoms.
(ii) The second number from the right represents the number of hydrogen
atoms +1.
(iii) The third number from the right represents the number of carbon atoms
—1.
(iv) Spare valencies are filled by chlorine atoms.
(v) The fourth number, if present, is the number of double bonds, and a prefix
C indicates that the compound is cyclic.
(vi) Conversely, if 90 is added to the code number, the last three digits, starting
from the right, give the number of fluorine, hydrogen and carbon atoms
respectively, with chlorine atoms to make up the valency requirements.
Details of molecular weights, vapour pressures and boiling points for the three
most commonly used chiorofluorocarbon propellants are given in Table 40.1.
Propellant 12 may be used on its own, without auxiliary pressure depressants,
as for feminine hygiene sprays. It is more usual to associate this propellant with
810 Harry's Cosmeticology
Vapour
Inter- Molec- pressure
national Chemical ular Bollin; at 21°C
Propellant Code No. formula weight point (psig) (kPa)
Dichlorodifluoromethane 12 CCl2F2 1209 —298°C 702 484
TrichlorofluOromethane 11 CCI 3F 137.4 —238°C 134 925
Dichlorotetrafluoroethane 114 C2 C1 2F4 1709 + 36°C 129 889
811
Aerosols
atmosphere is very much less complete than had been assumed. The atmosphere
is being revealed as a very complex interlinked system and the consideration of a
single effect or reaction sequence in isolation can give very misleading results.
Numerous reports, for example that by Brasseur, 12 have been written
reviewing the scientific developments together with calculations of potential
ozone depletion. Such calculations, which are based on many assumptions, may
be regarded as doubtful forecasts as they are a compound of uncertain science
and uncertain scenarios projected many decades ahead. Research into the
problem continues and new assessments of the science are prepared as more
information becomes available.
Statistical treatment of actual ozone measurement data has so far failed to
reveal a depletion trend.
Notwithstanding the lack of scientific validation of the Rowland—Molina
theory, some countries have regulated the use of chlorofluorocarbons as aerosol
propellants.
iJ,ydrocarbon Propellants
Hydrocarbons such as propane, n-butaneand isobutane, which have a low order
Of toxicity, 13 are used in aerosol packs as cheap, stable and non-corrosive
propellants which are handled safely in transport, storage and filling
operations. 116 They are used with water-based products to give low-cost
non-flammable end products that present no special corrosive hazards. They are
blended with other propellnts and solvents to lower the overall costs and in
some cases actually improve the product.
These hydrocarbons, which must be relatively free of impurities, may be
blended with each other or with the chlorofluorocarbons to give the required
vapour pressures for the propellant charge content to dispense aerosol products
satisfactorily.
Propane, n-butane and isobutane are the only naturally occurring hydrocar-
bons which have the stability and availability that make them suitable for use on
a commercial scale. These hydrocarbons are readily separated from naturally
occurring mixtures to give stable odour-free propellants in large volume at low
cost. Their physical properties are given in Table 40.3. Hydrocarbons are used'
alone in water-based aerosol products such as shaving creams, window cleaners,
812 Harry's Cosmeticology
Table 40.3 Physical Properties of Hydrocarbon Propellants
Boiling
point at Vapour pressure
Specific gravity 1 atm at 106F (37-8°C)
Propellant of liquid Formula M.W. (°C) (psig) (kPa)
Propane . 0.508 C3H5 44-09 —42•1 189-5 1306
n-Butane 0-584 C4H10 5812 —405 520 358
Isobutane 0-563 C4H,0 5812 —11-7 735 507
air fresheners and spray starches. They may also be used alone or combined with
other propellants to dispense many water-free products.
Hydrocarbons with their low specific gravity are lighter than most products so
the liquefied hydrocarbons will float on top of the product when they are
immiscible.
Herzka' 7 has indicated that there are advantages in using propellants that are
lighter than the product; in particular there is no danger of discharging pure
propellant with three-phase systems, at least until all the product has been
dispensed.
Besse, Haase and Johnsen' 8 have presented graphs showing that up to 30 mole
per cent of propane can be blended with propellant 12 before a pressure of 100
psig at 70°F is reached (689 kPa at 21.1°C).
Smaller amounts (about one-third by weight) of hydrocarbon propellants,
because of their lower specific gravity, are required.compared with chiorofluoro-
carbon propellants in order to give the same pressure in an aerosol container and
to give the same volume content in a formulation.
The hydrocarbon propellants are soluble in chlorofluorocarbons, alcohol,
chloroform, methylene chloride, ether and the higher hydrocarbons such as
n-pentane and n-hexape. They are non-polar and are not hydrolysed by water.
There is little chance of the formation of corrosive substances from the
hydrocarbons. This non-corrosive characteristic is an important advantage in the
use of hydrocarbons in aerosol propellants. They are also essentially immiscible
with water. Butane-water-alcohol systems, designed to produce a single liquid
phase, are of increasing importance for personal care products such as hair
sprays. Care needs to be exercised with respect to corrosion and therefore
product/container compatibility.
Hydrocarbon propellants have a low toxicity rating, and they do not form
toxic decomposition products at high temperatures. Normal care should be
exercised, however, to avoid contact of the liquid with the skin or eyes.
Hydrocarbons have low flash points and low explosion limits.
Flash point (°C)
Propane —104
Butane -74
Isobutanc -83
However, from the point of view of flammability the properties of the complete
formulation are most important. Many formulations using hydrocarbons as the
Aerosols 813
per cent
w/w
Propellant 11 45
Propellant 12 45
Isobutane 10
Similar blends are being brought into use increasingly for alcohol-based hair
sprays and deodorants.
Fowks2° cited an example in which the length of the flame in flame extension
tests was notably reduced when a chlotofluorocarbon propellant in a hair spray
was replaced by a chlorofluorocarbon–hydrocarbon mixture.
Calor Gas (Limited) UK supplies propane–butane mixtures now used as
propellants in a large number of important aerosol products. They offer
'destenched gas' as opposed to 'unstenched gas'—the latter being collected at
the refinery and containing a small quantity of residual sulphur compounds.
Destenched gas is obtained by further processing through molecular sieves or
other absorbents. Calor aerosol propellant is available as a standard at pressures
of 30, 40, and 48 psig (207, 276 and 331 kPa) at 21°C. The different pressures are
obtained by varying the percentages of butanes and propane present, since each
has a different vapour pressure.
Dirnethyl Ether
Rotheim, in Norway, first suggested the use of dimethyl ether for aerosols .21
Since 1966, aerosols using dimethyl ether as a propellant have been proLluced,
with4e major contribution from Holland.
Highly purified dimethyl ether is an attractive propellant. 22,23 It is virtually
odourless, has specific dissolving powers and, as it is miscible with water, is
capable of producing single liquid phase product systems (with ethanol, for
example) thus commending its application to various personal care products.
Current toxicological studies '24 wide in their scope, provide encouraging data
to date. The toxicity studies continue and further data are awaited with interest.
Flammability characteristics of dimethyl ether are such that when it is handled
with appropriate caution and care, both in production and in product formula
tions, commercial experience reveals that no unacceptable hazard is posed.
Increasingly dimethyl ether is being used both in Europe and now in America
for many types of aerosol product, including hair sprays, setting lotions,
deodorants, deo-colognes, perfumes and toilet waters.
814 Harry's Cosmeticology
li
0
I. 0 .-
. 0 0
U lz
VI U & .2 . .
C
U o6 ulu 00
I-
0 o.
C
EVI
VI
()
•0
I = .0
V 0
.0
Va
VI
0 C
0
I-
0
C
VI .2 . .. >. .9 C
. . ii E
VIVID U.
.0
U
Q
CI
I V -.0
VI U
. .
J9 E I. .
.0
cn
. .-..0
0
o -
•0
I II .0 0
0.•. 0 0
—
VI
C
VU
CC
.0
•0 .
II
C
of
0 oE C 0
0. t4
I I.
z
U.-.—
Id 0
0.;.-
C0 U 0
z
fl
I
C
0
0
I)
0.
U,
U
•0
I C
. 0.
I-
. . 0.
C
. .U U
V .0
.
.E8
I
—53
816 Harry's Cosmeticology
safely used with tinplate and aluminium cans. Filling presents difficulties, and
has to be carried out as a two-stage process which slows down production, or else
by special methods such as pre-saturation of the product with carbon dioxide.
However, new generation valves may be employed which avoid all of these
difficulties.
Both nitrous oxide and carbon dioxide have been used in conjunction with
chiorofluorocarbon for dispensing space sprays.
Another mixed propellant system, developed by the Dow Chemical Co., is
based on a mixture of a soluble compressed gas such as nitrous oxide with
chlorinated hydrocarbons such as methylene chloride or 1.1,1-trichloroethane.
This system is claimed to be cheaper than the chlorofluorocarbons which are
replaced either in part or totally.
Depending on the solvency of the mixture and nature of spray required, the
chlorinated hydrocarbon may be methylene chloride or 1,1,1-trichloroethane or
blends of the two in the range 1:2 to 2:1. Grades of 1,1,1-trichioroethane
developed for cold degreasing or vapour degreasing are generally applicable to
aerosol use, for example:
Genklene N } ICI Ltd
Genklene LV
Chiorothene NU
Chiorothene VG Dow Chemical Co.
Gelled Propellants
When a product is dispensed from a.çonventional aerosol container, some of the
liquid propellant is included in the spray. The additional droplets of liquid
propellant can cause non-uniform deposition of product, a more rapid change in
the characteristics of the spray during long bursts owing to cooling caused by
evaporation of propellant, and increased cooling effect on skin with such
products as antiperspirants and deodorants. It has been claimed 30 that these
undesirable effects are minimized by the use of gelled versions of any liquefied
propellants. The propellant mixes with the solution of the product in its dry
vaporized form, and additional liquid droplets are not introduced into the spray.
The initial spraying times of gelled propellants were claimed to be approxi-
mately one-third greater than those of conventional propellants, and the overall
dispensing capacity was said to be approximately 40 per cent greater than that
provided by the conventional types of propellant.
FILLING OF AEROSOLS
As the industry began, the most important method for filling aerosols with
liquefied propellants was cold filling. This method is still used today for some
specialized filling, particularly in the perfume and pharmaceutical industries.
A second important method—under-cup filling—became a dominant method
of production filling aerosols in the late 1960s and the 1970s in the United States,
using machines from Kartridg Pak, USA. It was also adopted, particularly for
large filling operations, in other countries.
With ever increasing importance, pressure filling is the most important
production method employed world-wide today. Eminently suitable for lique-
fied propellants of all types, it is now also accepted as a proven method for
818 Harry's Cosmeticology
compressed gases provided that suitable aerosol valve designs, originating in the
1970s and improved in the 1980s, are employed.
Cold Filling
In cold filling, the propellant is cooled to a low temperature at which it will
handle as a liquid, and is run through the open aperture of the aerosol container
to join chilled product concentrate already there. The valve assembly is then
çaled on to the container.
This method facilitates the rapid filling of propellant while at the same time
expelling air from the aerosol container. It cannot be used where the chilling
effect produced by the addition of the cold propellant would adversely affect the
product concentrate as, for example, with some emulsions and water-based
products.
The temperature of the contents in the sealed aerosol container soon returns
to normal, assisted by passage through the essential hot water test bath.
In spite of the simplicity of the cold filling method, both cital investment
costs and refrigeration running costs are high.
Under-cup Filling
All operations in under-cup filling are conducted at ambient temperatures.
Product is filled into the open-aperture container. The valve, complete with
actuator, where dimensions allow, is placed loosely in the one inch (25 . 4 mm)
aperture and the filling head is lowered to seal over and around the container
top.
The valve is lifted slightly, a vacuum is drawn in the container head space, and
propellant is injected between the valve and the can curl. Then the valve is
sealed into the container aperture and the filling head recedes.
Pressure Filling
With pressure filling, the propellant at room temperature is injected under
pressure through the aerosol valve itself (see Figure 40.4).
The sequence of operations entails the addition of the concentrate to the
empty container, displacement of air from the head space i sealing on the valve
assembly, metering the required quantity of propellant and injecting it under
pressure through the aerosol valve.
With cold filling, the introduction of propellant iuto the open container
displaces almost all the air in the head space. It is self-purging operation.
However, when pressure filling is used, there is a seal between the charging head
and the aerosol valve; air remaining inside the container could not escape and
would be compressed as the propellant was added. This air would contribute to
undesirable excess pressures in containers and the oxygen in the air would affect
corrosion performance. Therefore it is necessary to evacuate the head space by
vacuum closing the valve on to the container, or by using an appropriate method
of purging.
Pressure filling machines employ pneumatic or hydraulic systems to operate
the charging cylinder. The quantity of propellant is varied by the adjustment of
Aerosols 819
ton
Stem
r;Aoun• sket
aling
mpound
Housi
ring
Diptu]
the stroke of the charging piston. The filling head on the valve is opened
mechanically or by propellant pressure, or both,to permit propellant injection
to take place.
Rapid injection even when there is a high propellant charge is possible with
current valves, irrespective of specification.
Pressure Filling with Burton off. After the valve has been'sealed into the
Harry's Cosmeticology
container aperture, propellant is injected through and arciUnd the valve stem.
Then the actuator button is fitted.
Pressure Filling Around the Button. The button is in place on the valve which is
closed on to the container. The filling head makes a seal concentrically on the
cup boss, and the injection of the propellant into the container takes place
primarily between the stem and the hole in the valve cup boss.
Pressure Filling Through the Button. The button is provided with gassing holes
and the seal is made by the skirt on the button between the cup boss and the
filling head. The propellant then enters the container through the gassing holes
in the button and again primarily between the stem and the hole in the valve cup
boss Cups and stems may be chosen to achieve very fast production filling and
line speeds.
Again, pressure filling is equally suitable for all liquefied and compressed
gases.
Solvent Saturation
Solvent saturation is an alternative procedure for compressed gases which entails
the pre-saturation, in a pressure tower, of part of the solvent used in the liquid
product concentrate with a soluble compressed gas, to the desired pressure,
before combining it with the balance of the concentrate in the container. This
procedure offers an alternative to expensive gasser shakers and allows improved
filling rates, making it attractive compared with other methods of filling
compressed gases. This type of aerosol filling has been adopted by a number of
aerosol fillers. However, the most important development in aerosol filling of
compressed gases is their metered introduction into containers through the
valves, using pressures and equipment similar to those employed for liquefied
propellants—that is, impact gassing with the latest aerosol valves.
Check Weighing
In production it is customary to weigh a proportion of the filled aerosol
containers, in order to check the accuracy of the filling operation, and to ensure
that filled units correspond with requirements.
TWO-PHASE SYSTEMS
A large proportion of all aerosols are two-phase systems containing a liquid and
a vapour phase (Figure 40.5). When the actuator, or button, of the aerosol is
822 Harry's Cosrneticology
phase
Liquid phase
-:--:-:-:-:-:--:- product + propellant
pressed, the pressure inside the container causes the liquid to rise up the dip
tube, through the open valve, and out of the button into the atmosphere where
the lower pressure allows the immediate expansion and vaporization of the
propellant and transforms the product into a spray mist, jet or foam. The solvent
will also evaporate rapidly leaving behind the active constituents in • a finely
concentrated and dispersed form.
The majority of two-phase packs consist of a homogeneous solution of active
matter, solvent and liquid propellant as the liquid phase, with propellant as the
vapour phase. Compressed gases which may or may not be soluble in the liquid
phase may also be used. The size of the emerging product particles is largely
determined by the propellant/product ratio, and is influenced by the design of
the valve. Examples of products based on two-phase systems are hair sprays,
deodorants and colognes. Space sprays, air fresheners and insecticides of the
solvent-based type are also two-phase systens.
The solvents employed in two-phase systems must effect a complete solution
of all the active ingredients, for example hair spray resin in ethanol, so that the
liquid phase of the filled pack is completely homogeneous. In addition to
dissolving the active substance, solvents in two-phase systems must also be
miscible with the propellant. Common solvents with these properties, and which
show sufficiently low irritancy and toxicity when used topically or inhaled,
include water, ethanol, isopropanol, propylene glycol and hexylene glycol.
Othçr solvents include methylene chloride and 1,1,1-trichioroethane.
Space Sprays
Space sprays of the two-phase variety require high propellant/product ratios to
achieve optimum particle size. Usually they contain not less than 80 per cent of
Aerosols 823
(4)
per cent
WIW
Triethylene glycol 45
Propylene glycol 46
Ethanol 60
Perfume 02
Propellant 11 425
Propellant 12 425
Surface Sprays
When the product constitutes 20-75 per cent by weight of the pack' (with
appropriate propellant loadings to complement), the resulting particle size of the
sprayed product may be of the order of 50-200 m. This type of spray,
exemplified by hair sprays, is referred to as a surface spray because the particles
are of such a size that they cannot remain suspended in the air but deposit and
coalesce on the available surfaces to which they are applied. They are sometimes
referred to as wet sprays, and include body deodorants, colognes, perfumes and
suntan sprays.
These sprays are commonly of the two-phase type. Although alcohol-based
products can be produced with a liquefied propellant concentration as low as
25 per cent, or even 5 per cent in the case of compressed gases, the resulting
sprays are relatively coarse.
Other personal products may contain as much as 95-99 per cent propellant,
for example feminine hygiene sprays, so chosen, in combination with a
vapour-base tap valve, that the chilling/freezing effect is avoided.
824 Harry's Cosmeticology
Some surface spray products using two-phase systems are described below.
Hair Spray
Example 5 is a typical hair spray formula.
(5)
per cent
Resin 15
Soluble lanolin derivative 02
Isopropyl myristate 02
Perfume 02
Ethanol 32•9
Propellant 11 39•0
Propellant 12 260
Internal pressure at 20°C: approx. 30 psig (207 kPa).
Suitable container: tinplate.
Tinted hair sprays can be obtained by incorporating coloured aluminium powder
or soluble dyes into the formula given in example 5. Colours and concentrations
must be carefully selected to obtain -the desired colour effect and also to avoid
valve blockage in the case of solid colorants. Up to 4 per cent of solid material
can normally be tolerated but it is advisable to include agitator pellets to assist
dispersion.
Hair Dressings
Oil-based hair dressings can be formulated relatively easily. Products can range
from simple solutions in mineral oil to more complex preparations containing, in
addition, fatty esters, hair conditioning ingredients or aqueous alcohol.
Body Deodorants
Giacomo 3 ' has given various formulations. The bacteriostatic and bactericidal
compounds discussed in Chapter 35 can be incorporated in aerosol formulations,
for example deodorant colognes.
(6)
per cent
Perfume 0•7
Bactericide 03
Isopropyl myristate 1.0
Dipropylene glycol 30
Ethanol 65-0
Propellant 12 - 30-0
Internal pressure at 20°C: approx. 45 psig (310 kPa).
Suitable container: lacquered tinplate or lacquered aluminium.
Foams
The two-phase system can be used to produce transient coarse foams, using
liquid propellants at low concentrations, particularly in product systems that are
Aerosols 825
relatively viscous. Stable foams can be obtained using compressed gases that are
solu'ble in the product phase. These systems may yet be shown -to have
application for toilet preparations.
Unchanged Products
Two-phase systems may also include certain unchanged products, such as pastes
and creams dispensed by the use of an insoluble compressed gas such as nitrogen
as the propellant. Large product retentions have been experienced, especially
with viscous products. This particular difficulty may be overcome by employing a
device which prevents contact between the liquefied propellant used and the
product. This has been covered by a number of patents. 32-35
The two-phase system using nitrogen as the propellant provides the simplest
and cheapest method of dispensing products in an unchanged form. It is essential
that the product is viscous enough to be extruded in a convenient form, yet
sufficiently fluid to flow readily in the container. A rheopectic product can be
ideal for this form of dispensing.
The most important product of this type to date is toothpaste, but many other
products including liquid soaps, hand and hair creams, suntan creams and
antiperspirant creams can be dispensed. Formulae can be adapted readily from
standard conventional products.
THREE-PHASE SYSTEMS
Propel Propellant
vapou vapour
Produ
-4--
Propellant +
Liquid product
propellant
(a) (b) (c)
Figure 40.6 Three-phase aerosol systems
Liquid Sprays
Sprays can be obtained even though the liquid propellant and the liquid product
are present as two distinct layers. Sprays can also be obtained by emulsifying the
liquid product in a lipophilic external phase consisting entirely or partially, of
liquefied propellant. These systems have been described in Du Pont pub-
!ications 36 and also discussed by Root 37 and by Sanders. 38 Both space sprays
and Wet sprays can be obtained with this type of system which ' can be used for
suntan sprays, room deodorants and others.
Powder Spray
Another variant of the three-phase system is the pressurized powder aerosol
which gives a powder spray. The formulation of such sprays was discussed by
Armstrong .39 They Consist of the product in particulate form constituting the
solid phase, while the propellant forms both the liquid and the vapour phases.
The liquid phase normally consists of the pure propellant, but sometimes it may
also contain a small amount of dissolved lubricant to assist in the dispensing of
the product by aiding the passage of the powder through the valve orifices.
Sometimes the powder, for example talc, may have been treated with the
lubricant. Because of the relatively low concentration of powder which can be
incorporated (less than 20 per cent and more usually 10 per cent by weight) this
system does not provide a very economic method for dispensing body powders.
However, it can be used for dry shampoos 40 and talc and for dispensing
antiseptic powders and other speciality products in powder form. Before use, the
dispenser is shaken in order to ensure that the powder is uniformly dispersed
throughout the liquid phase.
Caking and agglomeration have presented many problems in powder packs.
In early sprays, clogging of valves occurred frequently, and dispensing was often
poor. There is a limit to the amount of powder that can be incorporated in an
aerosol pack, above which severe clogging of the valve and/or dispensing head is
experienced, resulting in either intermittent discharge or complete failure.
When the powder tends to aggregate, the addition of a suitable surface-active
agent may be indicated.
The following formula illustrates this type of preparation.
(8)
per cent
Talc 90
Lubricant 05
Bactericide 01
Isopropyl myristatc 02
Perfume 02
Propellant 11 450
Propellant 12 450
Pressure at 20°C: approx. 35 psig (241 kPa).
Suitable container: lacquered tinplate or lacquered aluminium.
Electrostatic Charge. In anticipation of, and attendant upon, the virtual ban on
chlorofluorocarbons in the USA, alternative propellants have been sought and
found for all products, including such personal case products as antiperspirants
and hair sprays. In general, hydrocarbons were adopted, with the exception of
carbon dioxide in the field of men's hair sprays. Adoption of hydrocarbons for
personal care products raised questions of flammability which were resolved by
careful formulation and, more importantly, by choice of valve specification.
However, one issue remains, that of electrostatic charge created when spraying
powders, as With aerosol antiperspirants and talc sprays.
Such phenomena are well known scientifically in industry. The creation of
such electrostatic charges on chiorofluorocarbon-propelled antiperspirants, pro-
duced in their thousands of millions since 1970, was never an issue. When
hydrocarbon was substituted for chlorofluorocarbon as the propellant for
powder sprays it was recognized that the electrostatic charge on the discharged
contents might spark and cause an explosion.
There has been much activity in relation to this problem. In friendly and
competitive vein, two groups, one in America 4 ' and one in Great Britain ' 42 have
sought to define and agree test methods. Codes of safe practice have been
devised and are generally available. Individual companies 43 have been searching
for quenching agents for incorporation into product formulations. Water is
good, for example, but is unacceptable in antiperspirant formulations because of
adverse effect on the product and problems of corrosion.
Foams
Foams are normally obtained by the use of a three-phase rather than a
two-phase system. Aerosol foams usually contain up to 15 per cent, usually 6-10
Aerosols 829
per cent, of liquefied propellant emulsified with an aqueous product in which a
fairly high amount of surfactant is present. The container is shaken, the vahe is
actuated, and the emulsion which is forced up the dip tube emerges through the
actuator. As soon as it reaches the atmosphere the dispersed propellant drops
will immediately vaporize, and, in so doing, transform the product into a thick
foam. Aerosols of this type are often fitted with a special foam head or spout.
Such systems are discussed in some of the references quoted earlier in this
chapter and in other Du Pont publications, and also under the appropriate
product headings elsewhere in this book. The formulation of cosmetic foam
preparations has been covered in a number of patents.553
Foam packs have been used to dispense shaving creams, shampoos and suntan
foams.
While the constituents of the aqueous solution or the primary emulsion in the
system, including the type and amount of surfactant, will undoubtedly influence
the physical and functional properties of the foam, there is no doubt that the
consistency of the foam, the rate of its formatiOn and its stability will also be
determined to a large extent by the type and the amount of propellant used, and
hence the vapour pressure prevailing in the system. For a given concentrate,
high vapour pressure propellants and high propellant charges tend to produce
stiff, dry and elastic foams with poor wetting and spreading properties. Medium
pressure propellants will produce a softer and wetter foam, while low pressure
propellants and low propellant concentrations will give rise to soft, wet, less
resilient, slowly forming foams. Interest in compressed gases for aerosol foams is
developing.
When oil-in-water emulsion systems are to be dispensed as foams, the
propellant vapour pressure will usually be between 20 and 50 psig (138-345 kPa)
at 20°C and the propellant concentration will be 5-25 per cent of the total
formulation, but usually towards the Lower end of that range.
This system provides a very valuable method for applying a small quantity of
cream to the skin and thus has found application also for hand creams, hand,
cleansers and suntan creams. It has application for liquid soaps, 54 shampoos and
hair condit'iThers. However, the most important product in this class, until now,
has been the aerosol shaving cream.
Shaving Creams
Aerosol shaving cream emerges from the nozzle of the container as a fully
developed foam. For this purpose, propellants with vapour pressures of at least
25 psig (172 kPa) at 20°C are employed. A 40/60 mixture of propellants 12 and
114 at a level of 7-10 per cent may be used to produce a vapour pressure of 40
psig (276 kPa) at 20°C. With vapour pressures above 40 psig at 21C the product
will expand rapidly inside the valve actuator and emerge from the container in
the form of a compact foam. The texture of the foam alters with increasing level
of propellant, becoming more rigid and dry. It is interesting to note that butane
foams are not flammable and have virtually replaced other propellants for these
formulations.
The formula given in example 9 illustrates the composition of a shaving
foam.
830 Harry's Cosmeticology
(9)
per cent
Stearic acid 5.9
Triethanolamine 31
Propylene glycol stearate 30
Lanolin 10
Glycerin 20
Laurie diethanolamide 20
Perfume 0•5
Water 82•5
92 parts packed with 8 parts of Butane 40.
Internal pressure at 20°C: approx. 40 psig (276 kPa).
Suitable container: lacquered tinplate or lacquered aluminium.
Shampoo
For aerosol shampoos a 'stream' foam, also referred to as a 'lazy' loam,' is
preferred. This is a soft type of foam which will spread easily but will not break
down too readily. For this purpose, low pressure propellants with vapour
pressures of 5-20 psig (34-138 kPa) are used, which will cause the product to
emerge initially as a liquid or semi-liquid stream, which gradually expands to a
foam. Normally about 40 per cent active matter is required. When using
non-soap detergents, corrosion problems may be encountered. -
The choice of propellants 12 and 114 in foam-type preparations is governed by
their stability in the presence of water as well as by the fact that when used in
combination they will produce any required vapour pressure. The foam pattern
depends on the type of valve and actuator employed as well as on the propellant
concentration, propellant type and the actual formulation. If the product used
for producing foam is in the form of an emulsion, it should not be too viscous,
otherwise it will not mix readily with the propellant and will subsequently
dispense unevenly with accompanying spurting and propellant loss. In the case
of an emulsion, foaming is also influenced by the type of emulsifying agent that is
used in the preparation of the primary emulsion; those of the oil-in-water type
are usually selected.
Water-based Systems
In a three-phase system comprising two immiscible liquid layers and containing a
water-based product, the product forms one of the liquid layers, and the
propellant forms the other as well as the vapour phase. Two possible situations
exist, which are illustrated diagrammatically in Figure 40.6 a and b(page 826).
In (a) the propellant is heavier than the product and forms the lower liquid
layer. This is generally the case with chlorofluorocarbon propellants. In (b) the
propellant is lighter than the product (as in the case when the forjier is a
hydrocarbon) and floats on top of the product. In either case (a) or (b) it is
essential to shake the aerosol thoroughly immediately before use. In case (a) this
is necessary to allow the two propellant phases to equilibriate so that the
pressure in the head space is maintained; this occurs automatically with case (b).
Also in case (a) the length of the dip tube requires careful adjustment to avoid
Aerosols 831
dispensing neat liquid propellant which in turn causes product residue problems,
through insufficient propellant.
The product is dispensed from the container when the valve is actuated.
Special actuators are required to assist break-up of the coarse product particles,
even though shaking, which causes entrainment of propellant in the product,
materially helps in breaking .up the product spray.
Care in the development of water-based products will resolve any problems
which may arise with regard to the efficiency of the spray, compatibility of
product and container in respect of corrosion, and correct balance of formula-
tion and valve specification to control flammability when hydrocarbon
propellants are used. As far as corrosion is concerned, with both chlorofluoro-
carbons and hydrocarbons full storage tests must be conducted with any
development formulations. Mention may also be made of an article by Geary,55
in which he discussed water-liquefied propellant-co-solvent systems designed to
assist in the formulation of low cost water-based spray products.
Hair Sprays
In the USA 67 basic hair spray formulations, using 25 per cent hydrocarbons and
75 per cent ethanol, are modified to accept (say) 8 per cent water or 10-15 per
cent methylene chloride to provide solvency for the resins. Since these products
consist of virtually 100 per cent flammables, choice of actuator and valve, even
to include vapour phase taps '68 is designed to produce good sprays 69 but with
attenuated discharge rates 70 in order to control flame extension 71 '72 and other
characteristics. Performance of these products is generally good; they appear to
have a longer pack life.
It should be noted that butane–water–alcohol systems will tolerate 8 per cent
water in a single liquid phase. Care must be exercised to ensure correct choice of
metal container specification in order to avoid corrosion problems .73
Carbon dioxide has been adopted in the USA as an alternative propellant for
some men's hair sprays.
In Europe 74 the cost advantages of cheaper propellants are of course
recognized, but their use is not so extensive because there are several important
differences compared with America:
(10)
per cent
Methylene chloride 35
Chiorofluorocarbon 20
Hydrocarbon 1
Ethanol-based product concentrate J
Deo-colognes
This European product '71-80 also alcohol-based, suffers the same reformula-
tion problems as hair sprays except that methylene chloride does not enter into
the formulation equation. Topped-up (carbon di oxide— chlorofluorocarbon),
hydrocarbon, hydrocarbon—water—alcohol, dimethyl ether—water and even
straight carbon dioxide propellant systems have been tried, or are currently
marketed while the availability of traditionally formulated products continues.
the characteristics of metals and their environment, for example by the use of
protective lacquer films and the inclusion of corrosion inhibitors in the formula-
tion, or by the modification of the latter.
While it is not proposed to give here a detailed exposition of corrosion theory,
it is necessary to refer to the causes of internal corrosion in aerosol containers
and to discusss ways of controlling them.
There are three main causes of internal corrosion in aerosols:
(i) Change of the propellant stability in the product environment.
(ii) Attack by the product.
(iii) Electrolytic interaction of dissimilar metals.
ethanol, and showed that appreciable corrosion of tin and iron did occur when
metal strips were refluxed with such a mixture. The mixture causing maximum
corrosion after 42 hours reflux was:
per cent
1,1,1-trichloroethane 30
Ethanol 55
Water 15
In practice, the use of 1,1,1-trichloroethane, especially inhibited grades,
together with anhydrous ethanol, should not present serious corrosion prob-
lems, but the introduction of even small quantities of water will increase the
chances of corrosion and thorough testing will be necessary.
Galvanic Action
In metal containers for aerosols more than one metal is often present, so that
apart from a purely chemical attack by the product on the metal of the container,
there are possibilities of galvanic corrosion, and several sites where it can occur.
The two-piece aluminium container may constitute a greater corrosion risk than
the one-piece monobloc container. Bimetallic interaction can occur in alumi-
nium containers with tinplate bases or tinplate valve cups, especially with
aggressive products and where any protective oxide or lacquer coating has been
mechanically damaged. The tinplate base in a two-piece container will usually be
cathodic to the aluminium body, but the polarity relationship will depend on the
product characteristics. Cathodic metal is not at risk in respect of corrosion but
forces metal which is more anodic to dissolve.
Even if all the container components are tinplate, two metals (tin andiron)
are still present. The electrical properties of these may depend on the oxygen
concentration. With some products when the oxygen concentration is low the
tin may be anodic with respect to the iron, but at higher oxygen concentrations
the tin may become cathodic and the iron anodic, and the latter becomes
corroded. With other products the whole pattern may be reversed. The optimum
level of air is a function of the product—package combination, and so vacuum
purging of tinplate containers may be effective in reducing the potential
corrosion of certain products.
Bimetallic corrosion can also be caused by traces of copper and by salts of
heavy metals which must therefore be minimized. The adoption of blackplate
and tin-free steel materials requires careful consideration.
adhesion to tinplate; because of this they are used as a base coat in many
systems. An internal resin stripe is present in many containers to protect the
product from discoloration or changes in odour or flavour following exposure to
the metal at the side seam. Side stripes are applied, usually by spraying,
immediately after the welding (soldering) operation, and the materials used for
this purpose include oleoresins and vinyl resins.
The internal lacquering of aluminium containers is different from and not as
easy as that of tinplate containers. While the latter are lacquered on the flat plate
before fabrication, aluminium containers must be lacquered by spraying or
flushing the finished container—a process which is rather expensive when
dealing with monobloc containers. It is preferable, however, not to rely on the
protection afforded by the lacquer coating, particularly where aluminium
containers are concerned, but rather to use, as far as possible, non-corrosive
formulations.
Anodizing
An alternative, or additional, method of protecting the internal surface of an
aluminium container is the process known as anodizing, whereby the naturally
occurring thin film of aluminium oxide is reinforced by a coherent, relatively
thick (4-6 m) and highly resistant layer of aluminium oxide. This film may be
coloured to give, for example, a golden shine effect. Anodizing is expensive but
gives a greatly increased corrosion resistance.
Electro-corrosivity86'87
Sophisticated and simple electrochemical methods for corrosion prediction have
been devised over the years. Their basic purpose is to study the relationship
between current and potential applied to all combinations of the basic metals
occurring in a particular metal container construction, in the product environ-
ment for which the specific container is intended. Account is taken of the areas
of metal exposed in the container as specified.
Significant advances in these techniques augur well for the aerosol industry in
the 1980s, promising even greater confidence than has already been well
established in making predictions about container behaviour after only 24 hours
of study. Increasing experience facilitates the elimination of unsatisfactory
specifications for the can, while screening out difficult product ingredients, for
example unsatisfactory perfumes or surfactants.
Rapid exploration of alternative materials such as corrosion inhibitors ensures
that only the product variations most likely to succeed in cans with favourable
specifications are put on full storage tests.
In the event of complaint or failure, the services offered by electro-corrosivity
facilities are invaluable in the inevitable investigation and especially when
marketing continuity demands a temporarily substituted but compatible pro-
duct.
Storage Tests
Full and adequate storage tests, whose design recognizes the aerosol product, its
application and the market for which it is intended, are vital. 88'89 Failure to
Aerosols
839
observe this simple rule is a recipe for ultimate disaster. Many filled aerosols
should be stored, upright and inverted, with air levels which reflect production
variations including lapses, both at ambient and elevated temperatures, for
periods commensurate with the expected shelf life.
Experience should be developed to acquire confidence that results obtained
after eight months at 35°-37°C, provided that the results at ambient tempera-
tures are also favourable, are a reasonable guide to performance over two years
in temperate climates. More rigorous testing is likely to be required for tropical
regions.
ALTERNATIVE SYSTEMS .
which operated on the bag-in-can principle, had thin aluminium foil in place of
plastic bags, suspended within a standard aerosol can. The mode of operation
was the same.
From Rhen AG, Presspack, and from Comes in Switzerland, commercially
successful plastic bag systems are available. Recent innovation permits around-
the-valve propellant injection, eliminating the need for propellant injection
through an orifice in the can base with subsequent plugging with a rubber
component. Claims for 98 per cent delivery are made for such systems, which
have considerable potential application 92 in the packaging of food and phar-
maceutical products and toothpastes within a wide range of viscosities, where
the problem of product—propellant incompatibility has imposed severe limita-
tions on the range of products which could be dispensed in this way. No doubt
other applications in the toiletry field are possible to extend the range, and a
claim has already been made for depilatories.
Compared with conventional systems, for which large amounts of propellant
are used to bring about the dispensing, propellant contents are low for all
bag-in-can and piston systems. While conventional valves and spray actuators
may be used with these systems, the use of mechanical break-up actuators is
necessary to achieve acceptable sprays where appropriate.
Tn -aerosol9'
The principle of product—propellant separation embodied in the bi-aerosol was
extended to the so-called tn-aerosol system which contained two internal
containers. This permitted mixing of two products to activate them at the time of
dispensing, for example mixing of an alkali base with stabilized hydrogen
peroxide for bleaching hair. The outflow was controlled by a two-channel valve
and the two products from the inner containers were then mixed in the required
ratio, for example by valve nozzles of different diameters in the two-valve
channels. The introduction of the tn-aerosol paved the way to co-dispensing,
and marketing, of two-component and multiple-component products in a single
pack.
Co-dispensing Valves
All major valve manufacturers have worked on the development of valves for
.93.94
co-dispensing. The early work was reviewed in Aerosol Age in 1968
The concept of co-dispensing, that is the dispensing of two or more products
from a single aerosol container, was originated by Du Pont who secured two
patents959 ' disclosing valve modifications which would allow co-dispensing of
two materials from a single aerosol container. Among the first valve manufactur-
ers to obtain a licence under the Du Pont patent was the Clayton Corporation,
USA, who developed the 'Clay-Twin' co-dispensing valve allowing the simul-
taneous dispensing of two different products in specified proportions through a
common orifice. The introduction of co-dispensing valves has made it possible to
dispense various products, including hot shaving creams and hair dye formula-
tions.
Hot Shave Creams. To dispense a hot shave cream, the product incorporating a
reducing agent is stored in a can, while hydrogen peroxide is placed in a
Aerosols 841
laminated plastic bag. When the valve is actuated, the shaving cream and
peroxide are dispensed and combined in a balanced proportion. This results in
an exothermic reaction between hydrogen peroxide and the reducing agent (for
example a mixture of potassium sulphite and potassium thiosulphate) and
heating of the shaving cream. The proportions in which the two reactants are
provided and combined are important. Serious problems, but different ones,
develop with an excess of either component; insufficient peroxide gives only a
lukewarm foam, whereas excess peroxide can lead to a liquid rather than a
cream product. According to Du Pont, foam temperatures as high as 80°C have
been produced with certain combinations of oxidizing and reducing agents.
The self-heating of shaving formulations based on the oxidation reduction
principle has also been the subject of a US paten t97 issued to the Gillette Co. in
which thiourea and substituted thiobarbiturjc acid derivatives were mentioned as
reducing agents.
Venturi Spraying
Reference has already been made to the fact that in systems such as the piston
container, the Sepro-can and the Presspack system, where product and pro-
pellant are separated from each other by a plastic or metallic surface, it is not
possible to effect satisfactory atomization. The attempts to combine separation
of product and propellant with adequate atomization have led to the develop-
ment of new systems in which the venturi principle has been utilized. The
prototype of this s y stem was a normal aerosol container which was filled with the
liquid propellant and connected by means of a plastic bridge with a glass vessel
holding the product to be sprayed. In Connection with these developments,
reference must be made to the Innovair system of Geigy, France, and the Preval
Atomizer developed by the Precision Valve Corporation of America. In both
these systems propellant and product are separated and only brought together at
the moment of spraying. There is no need for the container holding the product
to be pressure-resistant, which thus permits the use of materials other than
metals.
842 Harry's Cosmeticology
The innovair System (iTO). The Innovair system, originally knoin as ITO and
used for insecticides and later air fresheners, was developed by Geigy S.A. of
Switzerland and employed a blow-moulded non-pressurized container which
held the product to be sprayed. The technical bulletin 98 dealing with this system
described it as follows:
Several technical advantages were claimed for this system. The outer con-
tainer which is no longer under pressure can be made from a number of plastic
materials which in turn permits the use of containers in a variety of shapes and
colours. It also provides chemical and corrosion resistance, thus overcoming
many of the problems associated with the possible incompatibility in the pack of
product and propellant encountered in tinplate and in aluminium containers.
The system allows, for example, for the packing in aerosol containers of
water-based products without the danger of producing extensive corrosion and
can be used to dispense solutions, emulsions and suspensions. The unit can only
be used in the upright position.
As with co-dispensing valves, maintaining the design balance for all produc-
tion units, recognizing product tolerances, and all conditions and extremes of
usage render an excess of one component over the other inevitable.
The Preval Sprayer90 . The Preval system, developed by the Precision Valve
Corporation, USA, also makes use of an aerosol cartridge consisting of a valve
and a dip tube assembly, incorporating its own propellant chamber—again
making it possible to eliminate the use of pressurized containers for the product
and to use metal, glass or plastic. The difference between the Innovair and the
Preval systems lies in the fact that in the former the product is supplied to the
venturi nozzle around the propellant container, while in the Preval system
the product passes through the propellant container into the venturi nozzle.
Again, since the propellant joins the product at the point of discharge, the
majority of compatibility problems hitherto encountered will not arise. Also, the
Aerosols
843
use of the Preval cartridge will result in a reduction in the amount of propellant
used, permitting the use of much lower propellant/product ratios. The Preval
system is suitable, for example, for personal care products for salons, touch-up
paints in garages, etc., as well as for do-it-yourself and hobby applications.
Aquasol
Aquasol'°' was introduced to the aerosol industry at the Chemical Specialties
Manufacturers Association meeting in Chicago, in May 1977.
Hydroc
arbon-propelled aqueous-based aerosol product systems, featuring
vapour-phase tap valves, achieve remarkable results in terms of particle size and
efficiencies when used for insecticidal and air freshener applications. Topically,
however, such sprays are generally wet, while, in contrast, application for
personal care products demands a dry spray.
For reasons of economy, water finds favour in aerosol products especially
where the formulator—for example because of legislative restriction on the use
of chl
orofluorocarbons —has to consider alternative propellants such as hydro-
carbon and seeks a ready modification of the flammability characteristics of the
sprayed product. Aquasol was designed and offered to achieve dry sprays with
such systems consisting—to quote an Over-simplified example—of equal parts
of hydrocarbon, water and ethanol, using a special valve and actuator button.
The established and simple principle is to introduce propellant gas into the
liquid product stream by means of a vapour-phase tap, and this takes place in the
housing or body of the valve. In the original Aquasol system the interaction of
gas and liquid streams took place in the actuator button. A high velocity vortex
of gas was arranged to strike the liquid stream tangentially just prior to release
from the actuator button orifice. The violent swirl initiated production of very
small uniform particles which resulted in a dry spray. For various design reasons
this swirl feature—.
-the interaction of the gas stream tangentially with the liquid
stream—was later transferred to the base of the housing, thus permitting the use
of simpler valves and a simpler design of actuator button.
Aquasol valve production, with designs at different levels of discharge rates,
finds many applications for various types of product formulations beyond, but
including, the hydrocarbon, water and ethanol systems originally proposed.
After 30 years of steady technical evolution of pumps the signs are that in the
United States, and elsewhere, this particular t y pc .f propellant-free spray
system—the pump—is achieving consumer acceptai
846 Harry's Cosmeticology
REFERENCES
12. Brasseur, G'., Critical Analysis of Recent Reports on the Effect of Chlorofluorocar-
bonson Atmospheric Ozone, Eur 7067 EN, Commission of the European Commu-
nitis, 1980.
13., Sanders, P. A., Aerosol Age, 1979, 24(1), 24.
1.4. Ayland, J., Aerosol Age, 1978, 23(3), 40.
15. Ford, G. F., Aerosol Age, 1978, 23(3), 41.
46. Shaw, D'.,. Aerosol Age, 1978, 23(8), 30.
17. Hezka, A., Aerosol Age, 1960, 5(5), 72,
18. Besse, J. D., Haase, F. D. and Johnsen, M. A., Proc. 461h Mid Year Meeting
.CSMA, 1960, p. 56.
19. Reed ., W. 1-I., Soap Chem. Spec., 1956, 32(5), 197.
20. Fowks, M., Aerosol Age, 1960, 5(10), 100.
21. US Patent 1 800 156, Rotheim, E., 1931.
22. Bohnenn, L., Aerosol Rep., 1979, 18(12), 413.
23.. Bohnenn, L., Aerosol Age, 1981, 26(l), 26 and 26(2), 42.
24. Reuzel, P. G. J., Bruyntjes, J. P. and Beems, R. B., Aerosol Rep., 1981, 20(1),
23.
25. US Patent 34 894, Lynde, J. D., 1862.
26. US Patent 256 129, Dccastro, J. W., 1882.
27. Anon,, Modern Packaging, 1961, 34(6), 88.
28. Hoffman and Schwerdtel GrnbH, Munich, technical literature.
29. Anthony, T., Aerosol Age; 1967, 12(9), 31.
30. US Patent 3 461 079, Goldberg, 1. B., 12 August 1969.
31. Giacomo, V. Di, Am, Pcrfurn. Aro,nal., 1957, 69(5), 49.
32. British Patent 740 635, Taggart, R., 1955.
33. US Patent 2 689 150, Croce, S. M., 1954.
34. US Patent 2 772 922, Boyd, L. Q., 1956.
35. US Patent 2 794 579, McKernan, E. J., 1957.
36. Du Pont, Publication KTM, 21.
37. Root, M. J, Am Perfum. Aromat., 1958, 71(6), 63.
38, Sanders, P. A., J. Soc. cosmel. Chem., 1958, 9, 274.
39. Armstrong, G. L., Soap Chem. Spec., 1958, 34(12), 127.
40. Hauser, N., Aerosol Rep., 1978, 17(5), 130.
.41. Reusser. R. E., O'Shaughnessy, M. T. and Williams, R. P., Aerosol Age, 1979,
24(3), 17.
42. Greaves, J. R. and Makin, B., Aerosol Age, 1980, 25(2), 18.
43. Johnson, S. C., Aerosol Age, 1979, 24(10), 28.
44. Rubino, A., Siciliano, A. A. and Margres, J. J., Aerosol Age, 1978, 23(11), 22.
Aerosols 847
45. Greaves, J. R., Manuf. Chem., 1980, 51(12), 3.
46. Pengilly, R. W., Manuf. Chem., 1980, 51(7), 49.
47. Turner, K., Aerosol Rep., 1981, 20(4), 114.
48. Berres, C. R., Aerosol Age, 1979, 24(7), 32.
49. Whyte, D. E., Aerosol Age, 1979, 24(5), 29, and 24(6), 30.
50. Dixon, K., Aerosol Age, 1979, 24(4), 20.
51. British Patent 719 647, Colgate-Palmohve-Peet Co. Inc., 1954.
52. British Patent 748 411, Spitzer, J. 0., 1956.
53. British Patent 780 885, Innoxa '(England) Ltd, 1957.
54. Coup!añd, K. and Chester, J. F. L., Manuf, Chem., 1980, 51(10), 39.
55. Geary, D. C., Soap Chem. Spec., 1960, 36(3), 135.
56. Von Schweinichen, J. G., Aerosol Rep., 1981, 20(3), 77.
57. MacMillan, D. and Simpson, A., Manuf. Chem., 1980, 51(6), 45.
58. Hyland, J. 0., Manuf. Chem., 1981, 52(5), 51.
59 Kelly, S. W., Manuf. Chem., 1981, 52(1), 33.
60. Special Report: Aerosol Age, 1978, 23(8), 16.
61. Braune, B. V., Aerosol Rep., 1980, 19(9), 294.
62. Blakeway, J. and Salerno, M. S., Aerosol Rep., 1980, 19(10), 330.
63. Anon., Manuf. Chem., 1978, 49(6), 37.
64. Anon., Manuf, Chem., 1980, 5(7), 31.
65. BAMA Standard Test Methods, UK, 1981.
66. CSMA Aerosol Guide. USA, 1981.
67. Murphy, E. .1., Aerosol Age, 1981, 26(3), 20,
68. Sanders, P. A., Aerosol Age, 1978, 23(10), 38.
69. Kopenctz, A., Aerosol Rep., 1978, 17(10), 335.
70. Kinglake, V. Aerosol Age, 1978, 23(8), 24.
71. Tauscher, W Aerosol Rep., 1980, 19(12), 412.
72. Tauscher, W " Aerosol Rep., 1979, 18(2), 60.
73. Nowak, F. A " Koehler, F. T. and MlcchelIi, A. L.,Aero.tol Age, 1978, 23(9), 24.
74. Klenliewski, "A., Aerosol Rep., 1981, 20(1), 8.
75. Eisberg N., Manuf. Chem., 1980, 51(2), 30.
76. Johnsen, M. A., Aerosol Age, 1979, 24(6), 20.
77. Special Report, Aerosol Rep., 1978, 17(11), 410.
78. Schonfeld, H. W., Aerosol Age, 1979, 24(5), 36.
79. Schonfeld, H. W., Aerosol Rep., 1979, 18(1), 5.
80. Schonfeld, H. W., Aerosol Rep., 1981, 20(3), 94;
81. Sanders, P. A., Soap Chem. Spec., 1960, 36(7), 95.
82. Archer, W. L., Aerosol Age, 1967, 12(8), 16.
83. Downing, R. C. and Young, E. 0., Proc. sd. Sect. Toilet Goods Assoc., 1953, (19),
19.
84. Parmlee, H. M. and Downing, R. C., Soap Sanit. Chem., (Special issue, Official
Proceedings CSMA), Vol. XXVI-CSMA (2).
85. Herzka, A. and Pickthall, J., Pressurized Packaging (Aerosols), London, Butter-
worth, 1958.
86. Murphy, T. P. and Walpole, J. F., Aerosol Rep., 1972, 11(11), 52S..
87. Ziegler, H. K., Aerosol Age, 1980, 25(9), 23 and 25(10), 26.
88. British Aerosol Manufacturers Association Test Methods,
Aerosol Age, 1981,
26(2), 26.
89. Kleniewski, A., Aerosol Rep., 1979, 18(7/8), 235.
90. Meurerch, H., Aerosol Age, 1967, 12(10), 32.
91. Anon., Aerosol Rep., 1968, 7(6), 265.
92. Braune, B. V., Aerosol Rep., 1981, 20(5), 171.
93. Anon., Aerosol Age, 1968, 13(1), 19.
.55
848 Harry's Cosmeticology